Skip to main content

Full text of "Journal of the Society of Motion Picture Engineers"

See other formats

From the collection of the 
j d 

* n 





San Francisco, California 




Volume XXXVI January, 1941 



The New Walt Disney Studio 


Growing Pains W. DISNEY 30 

Pioneering in Talking Pictures L. DE FOREST 41 

Twentieth Century Camera and Accessories 


Problems in Television Image Resolution C. F. WOLCOTT 65 

Televising the National Political Conventions of 1940 

H. P. SEE 82 

Motion Picture Editing 


Current Literature 105 

Society Announcements 109 





Board of Editors 
A. C. DOWNES, Chairman 




Subscription to non-members, $8.00 per annum ; to members, $5.00 per annum, 
included in their annual membership dues; single copies, $1.00. A discount 
on subscription or single copies of 15 per cent is allowed to accredited agencies. 
Order from the Society of Motion Picture Engineers, Inc., 20th and Northampton 
Sts., Easton, Pa., or Hotel Pennsylvania, New York, N. Y. 
Published monthly at Easton, Pa., by the Society of Motion Picture Engineers. 

Publication Office, 20th & Northampton Sts., Easton, Pa. 

General and Editorial Office, Hotel Pennsylvania, New York, N. Y. 

West Coast Office, Suite 226, Equitable Bldg., Hollywood, Calif. 

Entered as second class matter January 15, 1930, at the Post Office at Easton, 
Pa., under the Act of March 3, 1879. Copyrighted, 1941, by the Society of 
Motion Picture Engineers, Inc. 

Papers appearing in this Journal may be reprinted, abstracted, or abridged 
provided credit is given to the Journal of the Society of Motion Picture Engineers 
and to the author, or authors, of the papers in question. Exact reference as to 
the volume, number, and page of the Journal must be given. The Society is not 
responsible for statements made by authors. 


** 'President: EMERY HUSE, 6706 Santa Monica Blvd., Hollywood, Calif. 
*" 'Past-President: E. ALLAN WILLIFORD, 30 East 42nd St., New York, N. Y. 
** Executive Vice-President: HERBERT GRIFFIN, 90 Gold St., New York, N. Y. 

^Engineering Vice-President: D. E. HYNDMAN, 350 Madison Ave., New York, 

N. Y. 
**Editorial Vice-P resident: ARTHUR C. DOWNES, Box 6087, Cleveland, Ohio. 

*Financial Vice-President: A.S. DICKINSON, 28 W. 44th St., New York, N. Y. 
** '-Convention Vice-President : W. C. KUNZMANN, Box 6087, Cleveland, Ohio. 

* Secretary: PAUL J. LARSEN, 44 Beverly Rd., Summit, N. J. 

*Tresurer: GEORGE FRIEDL, JR., 90 Gold St., New York, N. Y. 

**M. C. BATSEL, Front and Market Sts., Camden, N. J. 

*J. A. DUBRAY, 1801 Larchment Ave., Chicago, 111. 

*JOHN G. FRAYNE, 6601 Romaine St., Hollywood, Calif. 

*A. N. GOLDSMITH, 580 Fifth Ave., New York, N. Y. 

*ARTHUR C. HARDY, Massachusetts Institute of Technology, Cambridge, Mass. 
**L. L. RYDER, 5451 Marathon St., Hollywood, Calif. 

*T. E. SHEA, 195 Broadway, New York, N. Y. 

*R. D. STROCK, 35-11 35th St., Astoria, L. I., N. Y. 

*Term expires December 31, 1941. 
**Term expires December 31, 1942. 


Summary. This article describes the planning and construction of the world's 
largest animated motion picture studio, explaining the routine and the problems en- 
countered in production of these pictures, together with the facilities provided in the 
new plant for solving these problems. The studio is unique in that it was built from 
scratch on a vacant plot of ground with no restrictions by existing facilities to hamper 
planning. It is comprised of some twenty buildings on a 51-acre plot and includes, 
among other things, one of the most complete air-conditioning plants in the country. 

The striking success of Walt Disney's feature-length cartoon 
Snow White and the Seven Dwarfs proved conclusively that there was 
a definite demand for pictures of this type, and that, if they were 
pictures of quality, they would not only be accepted but could be 
produced at a profit. The recognition of this fact prompted the Dis- 
ney Studio to launch an augmented program of feature productions 
which resulted in the release of Pinocchio early this year and the 
introduction to the public this fall of a new musical feature called 
Fantasia. The year 1941 will see a further increase in the number 
of features released. 

Augmented production schedules bring with them problems of 
expansion. They require increased personnel, more space, enlarged 
facilities, and smoother production flow lines. During the era prior 
to the release of Snow White the studio averaged eighteen to twenty 
short subjects in work all the time, and it required seven to nine 
months from the beginning of actual production until the picture was 
shipped. With that experience on short subjects it must be clear that 
on a program of feature production it would be necessary to have a 
proportionate number of features in work, depending on the release 

If a program of one feature a year were contemplated, it would be 
necessary to have three features in work one starting, the second in 

* Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received Novem- 
ber 1, 1940. 

** Walt Disney Productions, Burbank, Calif. 


4 W. E. GARITY AND J. L. LEDEEN [J. S. M. P. E. 

animation, and the third in the finishing stages. This would result in 
a two-year production period for each picture. If the release schedule 
were to be increased to two a year, there would have to be six features 
in process of production at one time. The only way to reduce the 
amount of work in progress at any one time would be to shorten the 
production schedule per picture, and this can not be dictated arbi- 
trarily since the work involved is largely creative and can not be ac- 
celerated at will like other activities. The type of story, the type of 
characters, and the number of characters are all factors in determining 
the length of the production schedule. 

When the studio decided to launch its program of feature produc- 
tion, steps were taken immediately to enlarge the studio facilities. 
An attempt was made first to redesign the Hyperion Plant, but after 

FIG. 1. View of new Disney studio, from nearby hillside. 

considering the long period of construction that would be necessary 
to revamp the plant and all the attendant confusion and disruption 
of production that such construction work would cause, it was de- 
cided to abandon the old plant and make a fresh start in a new loca- 

After considerable study, the present site in Burbank at the edge 
of Griffith Park was selected. The plot consists of fifty-one acres of 
land and on it, without restrictions of any kind, the studio engineers 
designed a plant suitable for the specific and specialized needs of the 
organization. (Figs. 1, 2, and 3.) 


The general process of making animated cartoon pictures is now 
fairly well known to most people, but a brief review of some of the 

Jan., 1941] 


finer details may be helpful in giving a more complete understanding 
of the reasons behind some of the features designed into the new plant 
at Burbank. 

The first step in the production of a cartoon is the selection of a 
suitable story and development of the necessary characters. For 
this purpose the studio maintains a sizable staff of story experts whose 
function it is to review existing stories, write new ones, and do de- 
tailed story research work. Sketches are prepared to illustrate the 
action which is involved and the artists start developing their con- 
ceptions of the characters who take part. Characters in a cartoon 

FIG. 2. View down "Minnie Mouse Blvd." showing, on the left, 
Inking and Painting, Camera, Cutting, Shorts, and Live Action Build- 
ings, and on the right the Animation Building. 

frequently go through quite a process of evolution before the final con- 
ception of the character is reached. 

Once the story is settled and the characters defined, the picture 
is ready for animation. Each picture has a director assigned to it 
and his function is the coordination of all production details. 

The animation of the action is first rendered in pencil by a staff of 
animators, assistant animators, and "inbetweeners," who have vari- 
ous sequences of the picture assigned to them by the director. Any 
time during the course of production the animator is at liberty to send 
a series of drawings to the Test Camera Department where the 
drawings are photographed in continuity. The film is developed in 
negative form and the tests thus developed are spliced in a loop. 



This test-film and the drawings are then returned to the animator 
who is provided with a moviola on which he can run this test loop to 
check the animation and make whatever changes that occur to him. 
When these test loops have been approved by the animator, they 
are delivered to the Cutting Department where the test is cut in con- 
tinuity and synchronized with the sound-track of the production. 
Each director reviews such tests from each of his pictures once a 
week in a small review room equipped with sound. When all the test 
animation is complete so that a complete picture is available in test 
form, the test animation reels are previewed by the studio at large in 

FIG. 3. View of Dialog and Sound Effects Stage and Animation Building. 

a theater. It has been found from experience that the reaction of the 
studio group at these showings parallels very closely that of the av- 
erage theater audience. It is therefore possible in this manner to 
anticipate general audience reaction to pictures while they are still 
in the early stages of production, and to make whatever changes 
might be indicated without excessive expense. 

When the animator's drawings are completed and approved, they 
are delivered to the Inking and Painting Department where they are 
traced on celluloid sheets. Following the tracing of the outlines, the 
entire areas of the characters are painted in full color. This depart- 
ment is operated entirely by girls who perform all the operations in 
this stage of the process. Operating in conjunction with the Inking 


and Painting Department is the Paint Laboratory, where paints are 
mixed, studied, and classified; and the Process Laboratory, which 
at the present time is more or less of an experimental institution, 
whose function may be described best as the study of special proc- 
esses for eliminating "jitter" from the animated pictures. 

When the inking and painting of the celluloids has been completed, 
they are sent, together with the proper background paintings, to the 
Production Camera Department where each picture is photographed 
in proper relation to its background and in proper sequence for ani- 
mation on a reel of sensitive film which, when printed and run at 
proper speed through a projecting machine, will give the intended 
animation to all the characters. The exposed film is sent out for de- 
veloping and upon its return is turned over to the Cutting Depart- 
ment for insertion in the final master reel. 

Of course, this studio, like live-action studios, requires extensive 
facilities for recording sound. Practically all pictures require special 
sound effects of various sorts. All pictures have various musical 
themes running through them and contain dialog, vocal, and in- 
strumental sequences. For the recording of music and effects the 
studio needs and has adequate facilities in the form of recording 
stages, monitor booths, recording equipment, and a re-recording 


The old Disney plant at Hyperion grew "like Topsy" and many 
planned developments and improvements there were handicapped by 
unavailability of property when needed. So as a result there could 
be no well ordered program of expansion laid out at that site. At the 
new studio, the staff had the advantage of starting from scratch. 
There were no existing buildings or utilities to hamper the planning 
and in this respect the new plant is probably unique in the history of 
motion picture studios. A detailed study was made of the studio's 
needs, and then, practically without restriction, a plant was built 
to fill those needs. 

It might be said that the uppermost thought in the minds of those 
who designed the new studio was to produce facilities for optimum 
per capita production. The intent was not to design a plant at 
lowest possible initial cost, but to provide at reasonable total cost 
whatever facilities, space, and organization increased production 
would justify. Emphasis was placed on steps to restrict maintenance 

8 W. E. GARITY AND J. L. LEDEEN [J. S. M. P. E. 

cost and operating cost in the new plant, and wherever reasonable 
initial expenditure promised to curtail operating overhead, that 
expenditure was made. 

The first consideration was to provide a smooth and efficient pro- 
duction flow line a sort of picture assembly line. One building was 
set aside for the complete creative function from beginning to end. 
This building is properly named the Animation Building, and in it 
are the story departments, directors, layoutmen, animators, assis- 
tant animators, and inbetweeners, the latter three doing the main 
part of the job, the absolute creation of the animated picture. Across 
the street from the Animation Building, and connected with it by an 
underground all-weather passage, is the Inking and Painting Build- 
ing, from which branch the Paint Laboratory and the Process Labora- 
tory, housing activities that are supplemental to the Inking and 
Painting process. From the Inking and Painting Building the pro- 
duction flow line continues smoothly past a checking unit into the 
Camera Building, and just beyond the Camera Building, in logical 
progression as to function, lies the Cutting Building, which repre- 
sents the terminus of the process. 

Three recording stages are provided: one for orchestra, one for 
dialog, and a third for sound effects. These stages, together with a 
theater, are arranged so that they are adjacent to both the Animation 
Building and the Cutting Building. A large live-action stage is pro- 
vided for special studies, and a restaurant has also been provided. 

Three definite requirements for specific internal conditions in these 
buildings pointed to the need for an extensive air-conditioning plant 
on the lot. First was the need for scrupulous cleanliness in the build- 
ings. At the old studio considerable trouble had been experienced in 
the Inking and Painting, Camera, and Cutting Departments because 
of inadequate control over the cleanliness of air and quarters. When 
it is considered that a slight speck of dust under the glaring lights of 
the cameras can produce undesirable effects, the necessity for scrupu- 
lous cleanliness becomes apparent. In the new studio, it was planned 
to attack this problem from three angles: to use sealed windows and 
weather-stripped doors wherever possible to keep dirt out, to avoid 
drapes and carpets in those buildings where dust and lint were a 
problem, and to provide complete air-conditioning for the mainte- 
nance of clean air. 

The second problem at the old studio had been the lack of control 
over atmospheric humidity. Inasmuch as paints used in making 


these pictures are water paints, their rate of drying and the condi- 
tion of the pictures after the paints have dried are both very de- 
pendent on the humidity of the air. If the air is too dry, the paints 
crack and chip off the celluloid, and if the air is too humid, the paints 
become tacky and smear. In the past these problems have necessi- 
tated frequent reworking of scenes and adjustment of paint formulas. 
In the new studio it was planned to have controlled air-conditioning 
so as to permit the stabilization of paint formulas and the reduction 
of losses. 

The third requirement was the simple need for comfort. In the 
San Fernando Valley temperatures have been recorded as high as 
109 degrees, and when it is considered that approximately ninety 
per cent of the cost of a finished picture consists of labor, and that 
in this category about seventy per cent of the labor is exclusively 
creative, the vast importance of ideal working conditions is easily 
recognized, since to keep a man comfortable is to keep him productive. 

The importance of adequate light in any office building has always 
been appreciated, but in an office building which consists of over 
three hundred artists' studios adequate lighting is doubly important. 
With this in mind, the rooms in the Animation Building and Inking 
and Painting Building were planned so that every artist was in an 
outside room, that as many rooms as possible would have north 
light, giving a maximum of light without glare, and that all rooms 
facing in other directions would be equipped with adjustable shutter 
awnings admitting light from the sky but barring glare from the sun. 

In keeping with the concern for employees' comfort and adequate 
lighting, thought was given to the provision of exercise and recreation 
facilities. Plans in this respect are not yet entirely complete, but it 
may be mentioned that there have been provided on the roof of the 
Animation Building a gymnasium, sun-deck, steam room, massage 
room, and other facilities for the employees. Art is a sedentary 
occupation and does not provide much opportunity for exercise. A 
gymnasium which helps the men keep fit can not be considered in the 
light of a luxury, but must be regarded as a necessary production tool. 
Further extension of exercise and recreational facilities on the lot 
are planned at a later date. 

At the old studio at Hyperion there was a recognizable lack of a 
suitable theater or auditorium for reviewing animation test reels, 
previewing pictures, and holding special study groups. The sound 
stage at the Hyperion Studio was a sort of "catch-all" for sound 

10 W. E. GARITY AND J. L. LEDEEN [J. S. M. p. E. 

recording, theater functions, re-recording, and all other activities 
that required a large auditorium space. In planning the new studio 
it was decided to untangle these conflicting activities and build a 
separate theater, distinct from the sound stages. This would make 
it unnecessary to interfere with set-ups on the sound stages and 
would thus eliminate interferences with production, at the same time 
providing a review room and an excellent place for re-recording 
music, dialog, and sound effects on their final sound-tracks, since the 
re-recording mixers would hear the sound in an actual theater at the 
final re-recording. 

Past experience with the studio's activities had indicated the 
importance of flexible designs that could accommodate themselves 
to expansion. Starting from scratch, therefore, in the planning of 
all the structures, underground utilities, storm sewers, sanitary 
sewers, water facilities, electric facilities, etc., provision was made to 
accommodate a certain amount of predetermined expansion in all 
the buildings on the lot that were likely to expand, and also to allow 
for expansion of these facilities to areas of the lot yet unused. This 
meant, of course, that future needs had to be estimated, pipe sizes 
made adequate, streets planned, conduit and piping arranged for 
easy expansion without interfering with current operation, etc. It 
can be said that as far as it has been possible for the studio engineers 
to anticipate the future expansion, provision has been made to 
accommodate it. 

Especially careful consideration was given to providing facilities 
which would tend to reduce maintenance and operating cost of the 
plant. In general, as one makes machinery more accessible and 
automatic in its operation, one tends to reduce the amount of time 
that has to be spent in its care and attendance. In the air-condi- 
tioning plant alone on this lot there is a considerable investment in 
machinery in the form of fans, motors, compressors, automatic con- 
trols, boilers, heat exchangers, and other equipment of this type. 
Considerable care has been taken in the design of this plant to pro- 
vide adequate access doors, to make things easily removable for 
repair, to provide suitable indicating devices giving a quick check on 
plant operation, to furnish servicing facilities and working quarters 
that require minimum steps on the part of the operator and minimum 
time to get things done. This is true not only of the air-conditioning 
plant but of the electrical facilities, plumbing facilities, and fire- 
protection system. For the amount of equipment on this lot, the 


size of staff required to care for it is considered small, and the re- 
sulting operating and maintenance costs are correspondingly low. 


A primary concern in the Animation Building being the provision 
of adequate light for all artists, the building was arranged in the form 
of eight separate wings connected to a central corridor, oriented on 
a true North-South axis. Each wing is three stories high, and in 
every wing floor the offices are arranged on either side of a wing 
corridor in such fashion as to make every office an outside room. 
This arrangement gives a maximum of rooms with true north light. 
Rooms facing other directions of the compass are provided with a 
specially designed Venetian-type awning with counter-balanced 
blades remotely adjusted from inside, permitting the artists to set 
the blades for maximum light from the sky with elimination of glare 
from the sun. 

The arrangement of the building in eight separate wings served 
another purpose beside the provision of light. Southern California 
being in the so-called earthquake belt, and the Animation Building 
being a structure some 250 feet long, it was desirable to separate it 
structurally into several units which, though integrated into the 
structure as a whole from the functional standpoint, could weave 
harmlessly as separate structural units in the event of seismographic 
disturbance. Each wing has, therefore, been constructed as a sepa- 
rate structural unit, connected to the central section by a copper 
expansion joint. The central section itself, because of its length, 
has been divided into two distinct structural units, making a total 
of ten structural units in this one building. Incidentally, the division 
of the building in this fashion made it possible to provide, at the 
junction points of wings and central sections, eight vertical shafts 
extending the full height of the building, perfectly located to carry 
the extensive system of air ducts included in the air-conditioning 
plant, as well as the fire-protection piping, the air-conditioning con- 
trol tubing, and other facilities. 

The production flow line in the Animation Building starts logically 
at the third floor and works downward toward the ground. The 
third floor is occupied largely by the Story Department whose func- 
tion has been described previously, and the Character Model Depart- 
ment, whose function it is to develop story characters by sketches, 
paintings, and three-dimensional sculptured models. Also, on this 

12 W. E. GARITY AND J. L. LEDEEN [J. S. M. P. E. 

floor are two large review rooms, each accommodating about fifty 
people, used mainly for the study of pictures by large groups, and also 
for story conference purposes. It is here, too, that Walt Disney has 
his office, since he functions very actively in the conception and 
development of every picture that is turned out. 

On the second floor, which is devoted to the direction units, each 
wing floor houses two direction units. Each unit consists of a suite 
of four rooms, accommodating the director, assistant director, layout 
artist, and secretary. Each two direction units are provided with a 
small review room, accommodating about ten people, for the purpose 
of projecting and checking the animation tests and test reels. These 
smaller review rooms use moviola projectors in place of standard 
projector machines, the moviolas being located in a booth adjoining 
the review room, and operated by a remote console from the review 
room. These smaller review rooms, incidentally, are called "sweat- 
boxes," a name given to them at the Hyperion Plant in a day when 
the rooms used for these purposes were stuffy, unventilated rooms 
which, when crowded with a group of animators, were truly and 
literally "sweatboxes." Of course, in the present air-conditioned 
plant the term has lost its original significance, but nevertheless con- 
tinues to be used for sentimental reasons. 

The first floor of the building is devoted exclusively to the anima- 
tion function, each room being designed to accommodate three men. 
The animator usually occupies a room by himself, with assistant ani- 
mators and in-between artists occupying adjoining rooms. 

On each floor, in each wing, there is provided a reception desk im- 
mediately off the main corridor. At these reception desks are em- 
ployed girls who operate as secretaries to all the artists in that parti- 
cular wing. Each girl operates a small telephone switchboard, and 
monitors all calls being made. She controls the door to the wing in 
order that no one may interrupt the artists without first being an- 
nounced. She receives and checks all materials passing to and from 
the wing. It is also her function to maintain records of the movement 
of materials between departments, and to keep time records, which 
are later used in cost analysis and control of production. 

The basement of the building is devoted to service functions. It is 
here that the Test Camera Department is located, with all necessary 
facilities for producing quickly test-films of the animation. Here too 
one finds the central telephone exchange, electrical shop, furniture 
storage, and all air-conditioning equipment serving this building. 

Jan., 1941] 



Other facilities available in the building are a library, in which the 
studio endeavors to keep the latest works available on subjects of 
interest to the studio staff, and a coffee shop, which provides room 
service to the artists, as well as counter service. 

Careful thought was expended not only on the design of structure 
and arrangement of rooms, but also on the furnishings and interior 
decoration. All the furniture in this building was specially designed 
and built to meet the needs of the studio. Obviously, the require- 

FIG. 4. Some of the special desks developed for the use of the artists. 

ments of furniture in an animation studio are so specialized that 
nothing in the way of standard equipment available on the market is 
quite adequate to fill the need. It was found necessary, therefore, to 
make functional studies of an animator's activity, and to design fur- 
niture so arranged that the animator could perform his work with 
minimum waste time and motion. All furniture in the studio is mod- 
ern in design and tone (Fig. 4). 

Aside from functional furniture, efforts were made to keep the art- 
ists at ease and comfortable by providing their rooms, wherever pos- 
sible, with adequate carpeting, drapes, and interior decoration. The 

14 W. E. GARITY AND J. L. LEDEEN [J. S. M. P. E. 

entire interior of the building was painted in harmonizing pastel 
tones, designed to furnish restful atmosphere. The purpose of carpet- 
ing and drapes is not alone ornamental, but serves to impose quiet on 
the artists' rooms. To this end, also, all ceilings in the building have 
been finished with acoustic plaster, to give further deadening effect, 
and even the light switches on the wall are of the new silent type re- 
cently developed. 

Inside, the building is kept comfortable at all times by one of the 
most complete air-conditioning systems in the country. Air is brought 
into each room at the center of the ceiling through a specially de- 
signed combination air outlet and electric light fixture, designed es- 
pecially for this installation, and never used elsewhere before. This 
unit has been designed so as to permit independent regulation of air 
quantity and direction, and to spread the air uniformly in a thin dif- 
fusing sheet across the ceiling. It is a pleasure to be able to report that 
with this unit completely draftless distribution of air has truly been 

Removal of vitiated air in the room is accomplished through an- 
other unique air outlet which is built into a fixture accomplishing a 
dual function. Around the entire outside periphery of the building, 
on each floor, there has been installed a continuous air-electric base, 
approximately a foot high and four and a half inches deep. The lower 
part of this base is an air duct, broken up into separate sections for 
each room, having a stamped louvered face and connected through 
the floor into an exhaust air trunk. The upper part of the fixture is a 
dual electric conduit, continuous around the building, with a re- 
movable face permitting the introduction at any time of extra tele- 
phone lines, small power lines, compressed-air lines, or other facilities 
that might be required. Through the use of this conduit, it is pos- 
sible to introduce such facilities without the necessity of cutting walls 
or breaking building structure. With this same thought in mind, 
all ceilings in the building have furred spaces approximately two feet 
deep above them, providing a permanently accessible space for the 
servicing of air ducts, sprinkler lines, electric conduit, and other fa- 

The Animation Building, oriented as it is to all points of the com- 
pass, and containing as it does numerous special conference rooms, 
rooms with heat-generating machinery in them such as projection 
booths, and rooms that are sources of heat and odor, like the coffee 
shop, presented a rather difficult air-conditioning problem. To solve 


this problem it was felt necessary to do a very complete job of zoning 
the controls so as to make conditions in the building independent of 
the position of the sun and independent of the movement of large 
bodies of personnel from one room to another. 

To this end, the building has been zoned systematically and thor- 
oughly with respect to all these factors, resulting in a total of 163 
separate zones of control, so flexibly arranged that it is even possible 
to heat any af these zones at the same time that refrigeration is being 
provided for any other zone. Air temperatures in the building are 
never permitted to go below 74 F, which seems to be the temperature 
desired by most of the occupants of the building. As the outdoor tem- 
perature rises during hot weather, the inside temperature is auto- 
matically increased until it reaches a maximum of 78 at an outdoor 
temperature of 100 . Relative humidity in the building is maintained 
between 40 and 50 per cent. An item of incidental interest is the 
fact that the four projection booths in the building, housing con- 
siderable heat-generating equipment, are all completely air-condi- 
tioned for the comfort of the occupants. 

A most unusual feature of the air-conditioning system in the Ani- 
mation Building is the provision of 100 per cent conditioned outdoor 
air, without any recirculation whatsoever. Ordinarily a plant taking 
this amount of outdoor air without recirculation would be prohibi- 
tively expensive to operate. Fortunately, on the site of the new 
studio, there is available underground an almost limitless quantity of 
clear well water, which remains at a temperature of approximately 
67 F all year round. This underground sea of water has been tapped 
by two wells, each delivering 2600 gallons per minute, and the well 
water is employed in the air-conditioning plant for pre-cooling the 
air in summer, and pre-heating it in winter. The use of well water 
in this fashion greatly reduces both the initial cost and the operating 
cost of the air-conditioning plant, which is thus privileged to em- 
ploy outdoor air in such large quantities, and without penalty. 
After nine months' work in this finely ventilated and odorless building, 
the studio feels that the decision to ban recirculation of air was a 
happy one. 


The three primary considerations in the design of the Inkers' and 
Painters' building being light, comfort, and cleanliness, the Inking 
and Painting rooms have all been arranged so as to receive north 

16 W. E. GARITY AND J. L. LEDEEN [J. S. M. P. E. 

light only, and through large areas of window. Furthermore, all 
working rooms have linoleum floors and lintless interiors, and windows 
are sealed to prevent the entrance of dust during the occasional dust 
storms to which the valley is subjected. Each work room is sepa- 
rately air-conditioned, being provided with 100 per cent conditioned 
outdoor air without any recirculation whatsoever, resulting in a low 
odor level in spite of the great predominance of volatile paints and 
odorous chemicals in these rooms. Each work space is independently 
controlled as to temperature and humidity continuously, the tem- 
perature being maintained in the same brackets as in the Animation 
Building, and the humidity being maintained at a constant value 
of 50 per cent. 

As an adjunct to this building there is provided a lounge, private 
restaurant, and sun-deck, available only to the girls who work in 
this building, and designed to afford rest and relaxation from the 
meticulous and exacting work they do. Lounge and restaurant are 
completely air-conditioned. 

In addition to the actual working spaces where the Inking and 
Painting processes are carried out, the building also houses super- 
visory offices, checking rooms where the production output is ex- 
amined before being passed on to the Camera Building, and a Paint 
Laboratory which makes all production paints, experiments with new 
colors, and carries on miscellaneous paint research. This laboratory 
has already classified and cataloged over 2000 different colors used in 
the studio's work. 

The provision of pure north light for each of the eight work corri- 
dors necessitated the leaving of open areas between alternate pairs of 
corridors, and these areas have been utilized to provide pleasant 
landscaping and flower beds which lend a friendly aspect to all rooms 
which open on them. 


The Process Laboratory was one of the most difficult on the lot 
to design, inasmuch as its functions are principally experimentation 
in photography and film processing methods, with the result that the 
designers were confronted with the problem of designing a building 
for future production methods which at the time were only vaguely 

Both the piping and air-conditioning systems in this building are 
particularly complicated. There is not much that can be said about 


these matters except to point out that water is required in many parts 
of the building at several different temperatures simultaneously, thus 
introducing problems of refrigeration and heating. Distilled water 
is distributed throughout the building, making necessary the use of 
corrosion-proof piping, in this case aluminum. Chemicals are used in 
great profusion in the experimental processes, making necessary the 
employment of stainless steel and rubber in the tanks, piping, and 
pumps employed in this connection. The air-conditioning had to be 
arranged to provide for fume removal at points of maximum chemical 
concentration, with ducts constructed to resist corrosion. Lead-clad 
steel sheets were used to construct corrosion-resisting duct work, and 
fans handling corrosive fumes were painted internally with asphaltic 

The building had to be designed so as to permit expansion and pos- 
sible rearrangement in floor space and facilities, and to this end 
piping was designed so it could be extended easily. Fans were selected 
to handle increased air quantities if needed, and duct work was ar- 
ranged to accommodate additions in space served. 


One of the keynotes of design in these buildings was cleanliness, 
and therefore, as in the Inkers' and Painters' building, windows are 
sealed, floors are waxed, no carpeting or drapes are employed, and 
doors are weather-stripped as protection against occasional dust 
storms. Furthermore, both buildings are carried under a slight in- 
ternal air pressure, so that air leakage, if any, is outward rather than 

In the Camera Building an additional problem arose from heat 
generated by lamps on the cameras. At the present time the building 
contains two standard cranes and two multiple plane cranes. Each of 
the latter is served by a 75-kw d-c generator, driven by a 125-hp in- 
duction motor. With the possibility of anything up to 75 kw being 
dissipated within a camera room 25 feet wide by 28 feet long by 16 
feet high, it is clear that special provision had to be made to remove 
this heat. To accomplish this, special housings were designed for the 
lamps and so arranged as to take a constant stream of cool room air 
past the lamp bulbs and out an exhaust duct to atmosphere. This 
not only keeps the lamps cool and increases their life, but also de- 
creases the amount of refrigeration required for maintaining comfort 
in the building (Fig. 5). 

18 W. E. GARITY AND J. L. LEDEEN [J. S. M. P. E. 

A unique feature of the Camera Building is that both men and ma- 
terials entering the camera rooms are pre-cleaned. All persons enter- 
ing the building have to pass through a special de-dusting chamber 
in which they are exposed to air blasts from twenty separate nozzles, 
directed against the clothing in such direction and at such velocity 
as to remove the bulk of the lint and dust picked up from outdoors. 
Painted celluloids entering the building from the Inking and Paint- 
ing Department first enter a special cell-cleaning room where the 

FIG. 5. One of the cameras in the new Camera Building. 

"cells" are first treated to discharge the static electricity which tends 
to accumulate on them, and are then brushed to remove the dust 
which had adhered by virtue of the electrical charge carried by the 

Another feature of the air-conditioning system in the Camera 
Building is the employment of exhaust air from the building to cool 
the camera generators on its way to outdoors. This has been done 
by casting in the floor of the generator room a concrete tunnel con- 
nected at one end to the discharge from the main exhaust fan, and 
opening at various points along its length under the motors and 

Jan., 1941] 



generators, thus directing a blast of air against these machines. An 
atmospheric relief in the ceiling of the room allows the exhaust to 
pass to outdoors. 

The Cutting Building has few features of unusual interest. It is 
naturally a fireproof structure. In addition to the regular cutting 
and splicing rooms and supervisory offices, one room has been set 
aside for the running of sound-track. This room contains dummy 
sound heads, and all sound-tracks received from the laboratories 

FIG. 6. Scene on Sound Effects Stage. 

are set up and run in this room, the sound being piped from here to 
any one of a number of locations on the lot that have been set aside 
for listening. Circuits are provided so that this sound may be re- 
produced in any of the directors' rooms, in the theater, and in cer- 
tain rooms in the Story Department. By setting up to run sound- 
track in this manner, the studio is assured of maintaining sound- 
tracks in good physical condition, since it eliminates transportation 
and handling, and insures preservation of the film in a clean atmos- 
phere at constant temperature and humidity. 

20 W. E. GARITY AND J. L. LEDEEN [J. S. M. p. E. 

Both the Camera and Cutting Buildings are provided with 100 per 
cent conditioned outdoor air, with no recirculation. The absence of 
recirculation in these buildings is a fire and fume protection, as well as 
an addition to comfort. The Cutting Building, furthermore, has all 
supply and exhaust air grilles equipped with fire dampers and damper 
guides, built integrally with the grille frames and arranged with fusible 
links so that a fire in any room will immediately shut off the air 
supply and exhaust from that room. 


Because of proximity to the Union Airport in Burbank, it was 
felt necessary to take particular pains with the acoustic insulation 
of the stages in which dialog, music, and sound effects were to be re- 
corded. Accordingly, it was decided to adopt the principal of con- 
struction sometimes referred to as "building within a building." In 
other words, these stages are constructed with double walls, the inner 
wall being entirely separate from the outer, and the ceiling of the room 
resting on the inner wall structure and having no connection with 
the outer. Both walls and ceiling are naturally provided with ade- 
quate layers of both hard and soft insulating materials to give the 
desired noise reduction. Acoustic absorption of the walls in the Or- 
chestra Stage, for example, has been measured as 52 decibels for sound 
coming from a gasoline-engine-driven tractor, and 70 decibels for a 
standard automobile horn (Figs. 6 and 7). 

The Orchestra Stage has been provided at one end with a hardwood 
orchestra shell constructed with diverging convolutions of such shape 
as to direct sound from the shell toward the microphones. The re- 
verberation time on this stage is approximately one second, and is 
practically flat at all frequencies, the variation being less than 1 

Each of the three recording stages has its individual recording chan- 
nel immediately associated with it. 

Air-conditioning systems for the stages naturally had to conform 
to severe acoustic specifications laid down for the stages themselves. 
In order to keep the air-conditioning systems quiet, fans were selected 
to operate at low speeds, ducts were sized so that nowhere did ve- 
locities exceed 700 feet per minute, elbows were curved and sharp 
corners avoided, and insulating materials were employed strategically 
in sound-traps to prevent fan noises from passing into the stages and 
thus to microphones. 



The need for a theater has been discussed in an earlier section of 
this article. The structure set up on the Disney lot for this purpose is 
a 622-seat house of concrete and wood construction, housing, in addi- 
tion to the theater and projection booth, a re-recording room in which 
are set up the dummy sound-heads employed in the re-recording op- 
eration. Immediately adjoining the dummy room is located a re- 
corder used exclusively for the production of final sound-track nega- 

FIG. 7. Scene on Orchestra Stage, showing convolutions of orchestra shell 
designed to improve acoustics. 

In the center of the house is a console for use by the re-recording 
mixers, containing the necessary mixing panels and volume indicators. 
This console is built in such fashion that it does not interfere with the 
visibility of any seat in the house, and yet gives the mixers an op- 
portunity to sit in the center of the theater while listening to the 
sound coming from the speakers. The theater is built with walls of 
acoustic plaster arranged in panels so oriented as to control the re- 
flections of sound. This theater has somewhat less reverberation 
than the average theater, but was intentionally constructed so be- 
cause during the re-recording operation the house is practically 
empty, and it is necessary to have a less reverberant auditorium to 
compensate for the absorption normally introduced by an audience. 

22 W. E. GARITY AND J. L. LEDEEN [J. S. M. p. E. 

The dummy room in the theater is also piped into the same sound 
channels on the lot as the special dummy room in the Cutting Build- 
ing, thus permitting sound-tracks to be run in the theater and repro- 
duced in many other places on the lot. Furthermore, although the 
sound recorder in the theater is ordinarily the only one used in the re- 
recording operation, sound channels are installed connecting the 
recorders in the various sound-stages with the theater to permit 
special recording operations when desired. 


Due to the fact that inadequate eating facilities are available in the 
vicinity of the new studio, it was felt necessary to provide on the 
premises a suitable restaurant to accommodate the employees. The 
structure erected for this purpose is a frame and stucco building con- 
taining one large dining-room seating approximately 400 people, in 
which only table service is available, and a smaller room housing a 67- 
stool counter in which only counter service is available. A small 
short order kitchen is provided behind the counter, but in general 
most food is prepared in the main kitchen, which serves both dining- 

The kitchen is equipped with every facility to be found in a mod- 
ern restaurant kitchen, the desire being to provide for the employees 
high-class food at moderate prices, even though it proves necessary 
for the studio to absorb part of the expense needed to accomplish 
this. Refrigerated walk-in boxes are provided for meats, vegetables, 
and dairy products, and refrigerated reach-in boxes for salads, pas- 
tries, and other foods. 

Both dining-rooms are completely air-conditioned, using 100 per 
cent conditioned outdoor air with no recirculation. This policy has 
made possible, in the restaurant, the exhausting of conditioned air 
from the dining-rooms through the kitchen proper, the exhaust taking 
place through ample exhaust hoods located strategically over the 
ranges, coffee urns, bake ovens, dish-washing machine, and other 
sources of heat generation. All windows in the kitchen are sealed and 
all doors are kept closed, with the result that the kitchen is kept al- 
most as comfortable as the dining-rooms themselves. 


The design and construction of this studio was favored by the fact 
that starting from scratch the buildings, the facilities within them, 


and the utilities that serve them were all designed simultaneously, 
thus permitting adjustments and compromises to be made during the 
process of design and avoiding many problems and complications 

This was especially true of the air-conditioning plant, since the 
General Electric engineers who designed the air-conditioning and 
supervised its installation were privileged to work with the architects 
and structural engineers from the moment they first began to set 
plans on paper. Building structures were thus laid out to accommo- 
date air-conditioning needs, and air-conditioning designs were ad- 
justed to meet structural requirements, with the final result that the 
plant as a whole represents a well integrated and coordinated entity. 

As a starting point in the design of the air-conditioning plant, 
buildings on the lot were divided into two groups those that have 
continuous usage, such as the Animation Building, Inkers' and 
Painters' Building, Camera Building, and so forth; and those build- 
ings which have intermittent usage, such as the theaters, stages, and 
restaurant. Stages may sometimes be shut down for weeks at a time, 
thereby not requiring air-conditioning during that period. The build- 
ings in the first group, however, operate every day of the week and 
require continuous service from the air-conditioning plant. It was 
decided, therefore, to serve the buildings in the first group (continu- 
ous usage) from a central heating and refrigerating plant, whose heat- 
ing and cooling facilities would be piped through water mains to the 
buildings in this group. The second group of buildings would be 
served by individual plants located within the buildings proper, hav- 
ing no connection with the Central Plant, so as to facilitate the shut- 
ting down of these buildings at any time desired without compro- 
mising the operation of the Central Plant. 

The Central Plant building, located between the Inkers' and 
Painters' Building and the Camera Building and slightly to their 
rear, lies very close to the center of the load it serves and is a rein- 
forced concrete structure, 40 X 175 feet, two stories high. The rear 
of the building is a large boiler room, and the front of the building a 
compressor or refrigeration room. In the center of the building, 
athwart the cooling and heating plants, lies the operating engineer's 
office, provided with double glass panels on all sides so as to permit a 
view of practically all the equipment in the building from the quiet 
interior of the office. In this office are kept various indicating and 
recording devices showing temperatures of heated and refrigerated 

24 W. E. GARITY AND J. L. LEDEEN [J. S. M. P. E. 

water, steam consumption, outdoor wet and dry-bulb temperatures, 
and other such items of interest in plant supervision. Also mounted 
in this office is an indicating panel of colored lights wired to show the 
operation of the 135 fans in the various remote fan rooms on the lot, 
and providing telephone circuits to permit two-way conversation be- 
tween the office and any fan room. 

The boiler room contains two 190-hp water-tube boilers, generating 
steam at twelve pounds' pressure. These boilers are normally gas- 
fired under completely automatic control, with standby equipment 
for burning oil under manual control. Part of the steam produced is 
passed into heat exchangers, where the heat is imparted to the water 
in a circulating loop which provides heat for all the buildings served 
by the Central Plant. The balance of the steam produced is piped 
directly to the fan rooms where it is injected through perforated cop- 
per tubes under automatic control into the air streams, to give regula- 
tion of humidity at such times as needed. With the large quantities 
of outdoor air employed in this air-conditioning plant, and the ex- 
treme dryness which the outdoor air reaches on many occasions (15 
grains of moisture per pound of air) the humidification requirements 
on the lot are approximately 5500 pounds of steam per hour at peak 

The evaporation of this quantity of water in any kind of humidifier 
would create a problem of disposal of salts of evaporation. In this 
plant the problem is solved simply by the use of water-tube boilers in 
conjunction with water filter and a Zeolite softener, together with a 
feed-water heating de-aerator. Both the liming of boilers and oxygen 
corrosion are thus prevented, and at the same time, having eliminated 
the necessity for use of boiler compound, the steam is kept absolutely 
odorless because it is pure and uncontaminated by foreign sub- 
stances. Regular blowdown of boilers keeps down salt concentration 
in tubes and drums. 

The boiler accessories, such as boiler feed pumps, condensate return 
pump, induced draft fans, and master control, are all electrically 
driven. The boilers are equipped with indicating pressure gauges, 
steam flow recorders, indicating draft gauges, and carbon dioxide and 
stack temperature recorders. 

The compressor room really contains two separate refrigerating 
plants, one of them serving the Animation Building, and the other 
serving Film Row, as the balance of the group of buildings served by 
the Central Plant is termed. The reason for segregating these two 


refrigerating plants is that air-conditioning for the Animation Build- 
ing is shut down every evening and started every morning, whereas 
air-conditioning for Film Row, being required to preserve paints in 
good condition, is operated on a 24-hour day, 365-day a year basis. 
It should be noted, however, that whereas these two plants are sepa- 
rated for normal operation, cross-over connections are provided so 
they may be tied together in case of emergency. 

The refrigeration plant consists of twelve 50-hp General Electric 
compressors, seven of them being connected to the Animation Build- 
ing and five of them serving Film Row. Each compressor is connected 
to an individual Struthers-Wells water chiller, and the chillers 
in turn are manifolded together into a common chilled water system. 
The refrigeration circuit of each machine is absolutely separate from 
that of any other machine, thus limiting the danger of losing refriger- 
ant through any leak which might develop. The refrigerant used is 
Freon-iJ?, and the total amount required in this plant is approximately 
2400 pounds, 200 pounds being needed for each machine. 

Control of the compressors is fully automatic and is regulated from 
outgoing chilled water temperature, the compressors being regu- 
lated so as to turn themselves off or start themselves up automati- 
cally in order to maintain a constant predetermined chilled water tem- 
perature. Capacity of the plant is approximately 550 tons of re- 
frigeration, or enough to cool 1650 gallons of water per minute from 
55 to 47 F, which are the operating temperatures on the chilled 
water loop. Two 20-hp pumps circulate the chilled water through 
the Animation Building circuit, and two 15-hp pumps operate on 
Film-Row. At times during night operation one 50-hp unit is ade- 
quate for the dehumidifying requirements of Film Row. 

Additional space is available in the Central Plant to accommodate 
any enlargement of plant capacity that might be dictated by expan- 
sion of buildings served from this plant. Piping leaving the Central 
Plant radiates in three directions. Running west, it crosses through 
a tunnel under the street to the Animation Building 250 feet away. 
Running north, it passes through a second tunnel into the basements 
of the Inking and Painting Building, Paint Laboratory, and Process 
Laboratory, which are all interconnected. Running south, it enters a 
third tunnel underneath the Camera and Cutting Buildings, and 
stubs off in such fashion that the Animation Shorts Building and 
Annex, moved over from the old studio, can be connected up to this 

26 W. E. GARITY AND J. L. LEDEEN [J. S. M. P. E. 

In the Animation Building, which represents about 350 different 
rooms covering 150,000 sq-ft of floor space, conditioning of the air 
is accomplished in eleven different fan rooms. Eight of these fan 
rooms are somewhat identical in arrangement, each of them serving 
all three floors on one wing. Two of the remaining fan rooms are 
smaller in size and designed to accommodate rooms in the basement 
which are set aside for test camera activities, the telephone office, and 
the production cost department. The eleventh fan room is on the 
roof of the building, and serves the gymnasium quarters in the pent- 

Inasmuch as one of the outstanding features of the air-conditioning 
plant is the zoning arrangement in the Animation Building, a brief 
description of a main fan room layout may be of interest. Outdoor 
air is taken into each wing separately through fresh air intakes lo- 
cated in the courts between wings, at second and third floor levels. 
This air is carried through a vertical shaft into the basement and first 
passed through a bank of air filters of permanent type, constructed 
of crimped galvanized wire arranged in sections of gradually decreas- 
ing aperture, and kept oiled with pure mineral oil of sufficient vis- 
cosity to scrub from the air all dust and pollen particles and other 
foreign bodies. Leaving the filters, the air enters the heat transfer 
surface of the well water coils which have a capacity in summer of 
cooling 100 air down to 75F, and a capacity in winter of heating 
28 air up to 55F. Control of the flow of well water in each fan 
room is automatically regulated from outdoor temperature to give 
optimum performance at minimum pumping cost. 

Next step in the conditioning process is humidification, and for 
this purpose perforated copper pipes extend into the air plenum, 
bringing pure steam reduced to two pounds' pressure from copper 
steam mains running from the Central Plant. Steam is ad- 
mitted through an automatic regulating valve controlled by a duct 
humidistat inserted in the air stream leaving the offices and exhaust- 
ing to outdoors. This part of the system insures that humidity in 
the building will never fall below 40 per cent, even during the ex- 
treme dry spells which occasionally reduce outdoor humidity to 
levels below 10 per cent. 

From the humidifier, the air is drawn into the main supply fan, 
which then pushes it into a plenum, branching at its further extremity 
into two separate plenums. At the head of one of these plenums is 
erected a chilled water heat transfer surface, and at the head of the 


other plenum is erected a hot water heat transfer surface, both of these 
coil banks receiving their water from separate piping systems ema- 
nating from the Central Plant. The need for separate piping for hot 
and cold water is occasioned by the necessity of being able to heat 
certain parts of the building at the same time as other parts of the 
building are being cooled. In this fashion, it will be seen that there 
is created at all times one plenum from which warmed air may be 
taken, and a second plenum from which cooled air may be taken. 
Each zone in the building is connected to both these plenums by a 
duct which travels from each zone through the furred ceilings down 
a vertical duct shaft and into the fan room, where it branches into a 
dual connection to the two plenums. A volume damper is installed 
in each branch and operated by an automatic damper motor which 
receives its control impulses from a remote thermostat located in the 
control room of the zone in question. This thermostat positions the 
two damper motors in such fashion that a proper blend of warmed 
and cooled air is taken from the two plenums and supplied to the 
rooms to hold proper room temperature. The action of the thermo- 
stat and mixing dampers is completely automatic at all times of the 
year, and no manual selection switch or change-over thermostat is 
required to decide whether the cooling or heating function is needed. 

Each of the fan rooms has approximately 20 zones of control con- 
nected to it. In general, each floor has been zoned separately, and 
only rooms having the same compass exposure are tied together on a 
single zone. Corner rooms which have double exposures are always 
on a zone by themselves, as are interior rooms which have no outside 
exposures at all, and conference rooms or other quarters where large 
numbers of people may gather suddenly for study or discussion. 

All rooms in each wing have their own exhaust ducts, connecting 
into a common exhaust trunk beneath the floor extending clear to the 
basement, where an exhaust blower picks up the vitiated air and 
blows it through the unexcavated portions of the structure past 
louvered screens and thus to outdoors. This accomplishes two inci- 
dental purposes : first, the forced ventilation of under-floor areas is a 
strong deterrent to the activities of termites, which might otherwise 
get into the building timbers; and second, the warm air emerging 
from the exhaust louvers at ground level helps protect the flowers 
from early morning frost in winter. 

Air-conditioning systems in other buildings differ in detail from that 
in the Animation Building, but in general adhere to the principle of 

28 W. E. GARITY AND J. L. LEDEEN [J. S. M. P. E. 

completely automatic control with no recirculation of air. There 
were many special problems for which interesting solutions were de- 
vised, but these are of interest mainly to the air-conditioning techni- 

This system of mixing damper control, which has been used else- 
where before, but probably never on the scale shown in this building, 
has worked with excellent success. The building has now been in op- 
eration for some nine months, and, considering the conditions of oc- 
cupancy, widely fluctuating outdoor temperatures, and the large 
amount of glass area in the building exposed to the sun, there have 
been strikingly few complaints about atmospheric conditions. 

Air handled on the lot by the various air-conditioning units is ap- 
proximately 40,000,000 cu. ft of air per hour. This is forced through 
the various duct systems by 135 multi-blade centrifugal fans driven 
by electric motors totaling 275 hp. The duct systems through which 
this air is distributed required about 600,000 pounds of galvanized 
sheet iron and structural iron for their fabrication, and the 250,000 
gallons of heated or refrigerated water per hour needed to condition 
this air is forced through approximately three miles of piping by 18 
pumps, totaling 121 motor hp. 

The total air-conditioned floor space in the studio comprises about 
250,000 sq ft, which is roughly equivalent to a 12-story department 
store having a 130-foot frontage and 160-foot depth. All told, more 
than 1300 motor hp are required for all the air-conditioning functions. 


Well water is used on the new lot not only for air-conditioning, but 
also for lawn sprinkling, toilets, and special process uses in the Process 
Laboratory. Water for these purposes is drawn from two wells, each 
300 feet deep, and each provided with a 150-hp submersible pump 
and motor, sunk to a depth of approximately 80 feet below ground, or 
60 feet below the normal water table. 

These two pumps tie into a common underground piping system, 
into which there is connected an elevated surge tank which serves to 
even out the operation of the pumps. Control of the pumps is regu- 
lated from the changing water level in the surge tank, which actuates 
a controller whose function it is to run one pump or both pumps, de- 
pending on the level of the water in the tank. 

The amount of water set aside for the precooling function in the 
air-conditioning plant is 3000 gallons per minute at peak load, rep- 
resenting a cooling capacity of 1000 tons of ice per 24 hours. 



The new studio, consisting as it does of some twenty separate 
buildings arranged over an area of 51 acres, is in reality a small city, 
and had to be provided with many of the underground utilities with 
which modern cities are equipped. Thus, the studio had to con- 
struct for itself its own streets, storm drains, sanitary sewer system, 
sanitary water lines, underground fire-protection piping, fire hy- 
drants, a private telephone exchange, and a complete electric dis- 
tribution system. In addition, there is underground conduit to ac- 
commodate a studio-wide public address system, which is monitored 
from the telephone office, and a supervisory fire alarm system to give 
an immediate indication at a central point of the passage of water 
through any of the main sprinkler valves on the lot. This latter sys- 
tem is not yet completed, but complete provision has been made to 
accommodate it. 

Electric power enters the lot from the lines of the City of Burbank 
in a 4330-volt feeder which is carried into a main transformer vault 
through suitable oil-immersed circuit breakers for distribution to the 
lot. A central switchboard controls the branch feeders which radi- 
ate from this central point. Power is distributed around the lot 
at 4330 volts to three separate transformer vaults located strategi- 
cally at load center points. In these vaults the voltage is reduced 
through suitable transformer banks to 440 volts, 3-phase, 60 cycles 
to provide power for the operation of electric motors, and by means 
of other transformer banks is converted into 3-wire, 110-220-volt 
circuits for handling the lighting system. All motors of y 8 hp and 
larger are operated on 440 volts. The small fractional hp motors are 
run on 220 volts, single phase. 

Should the plant requirements ever exceed the capacity of the pres- 
ent main vault equipment, space has been set aside at the rear of the 
lot for erection of a studio sub-station capable of taking 33,000-volt 
service from the City of Burbank and converting it to the studio 

In addition to the regular lighting circuits, an emergency lighting 
circuit has been provided on the lot, servicing certain strategically 
located lights in corridors, boiler room, and other vital points, this 
circuit taking its energy from an automatic 25-kw gasoline engine- 
generator which starts up and reaches full speed within three seconds 
from the time of power failure. 



In the JOURNAL for December, 1938, there appeared a most inter- 
esting paper written by Dr. H. T. Kalmus describing the adventures 
of Technicolor in Hollywood. I have been asked to prepare an article 
along similar lines telling of highlights in the history of our company 
and animated pictures. Messrs. Garity and Ledeen have written a 
paper for the JOURNAL covering the technical side of our develop- 
ment, so I had better stay on my side of the fence and talk about 
animation and where I was born and about Three Little Pigs and 
what-about-the-future of the business. When I protested that all 
this had been written up too many times before, and that such an 
article would be dull and of little interest to the members of the 
Society, Mr. Garity said, "That's right!" and left the office with a 
dirty laugh. 

Making this job even more difficult, I found in rereading Dr. 
Kalmus' paper of 1938, that he had "lifted" semi-philosophic 
thoughts which I had planned to put in my article. I accuse him of 
what might be called "prophetic plagiarism," and I resent it, too, 
because I have so few semi-philosophic thoughts. 

For instance, Dr. Kalmus starts off by stating that his develop- 
ments in Technicolor have been an adventure, and adds the Webster 
definitions of adventure: chance of danger or loss; the encounter of 
risks; a bold undertaking; a remarkable experience, a stirring incident; 
a mercantile or speculative enterprise of hazard. Now, I had planned to 
start my paper with this definition and continue with the statement, 
"My business has been a thrilling adventure, an unending voyage of 
discovery and exploration in the realms of color, sound, and motion." 
It has been that! And it has been a lot of fun and a lot of headache. 
The suspense has been continuous and sometimes awful. In fact, 
life might seem rather dull without our annual crisis. But after all, 

* Received Dec. 1, 1940. 
** Walt Disney Studios, Burbank, Calif. 


it is stress and challenge and necessity that make an artist grow 
and outdo himself. My men have had plenty of all three to keep them 
on their toes. But how very fortunate we are, as artists, to have a 
medium whose potential limits are still far off in the future; a medium 
of entertainment where, theoretically at least, the only limit is the 
imagination of the artist. As for the past, the only important con- 
clusions that I can draw from it are that the public will pay for 
quality, and the unseen future will take care of itself if one just keeps 
growing up a little every day. 

The span of twelve years between Steamboat Willie, the first 
Mickey with sound, and Fantasia, is the bridge between primitive and 
modern animated pictures. No genius built this bridge. It was built 
by hard work and enthusiasm, integrity of purpose, a devotion to our 
medium, confidence in its future, and, above all, by a steady day-by- 
day growth in which we all simply studied our trade and learned. 

I came to Hollywood broke in 1923, and my brother Roy staked 
me to a couple of hundred. We lived in one room and Roy did the 
cooking. He was my business manager, and I didn't have any busi- 
ness. His job was to scare up three meals a day, and his job now is 
to conjure up three million dollars to meet the annual payroll. Both 
jobs have demanded just about the same amount of sweat, ingenuity, 
and magic. The main difference is that Roy sweats more red ink now. 
But no matter what the future deals me, I shall consider that I have 
come a long way, if for no other reason than that Roy dosen't do the 
cooking any more. 

I sold my first animated cartoon for thirty cents a foot. Pinocchio 
and Fantasia cost around three hundred dollars a foot. The first 
Mickey Mouse was made by twelve people after hours in a garage. 
About twelve hundred people are working overtime now in a fifty- 
one-acre plant with fourteen buildings, four restaurants, its own water 
system, air-conditioning, and a gentleman named Myron to massage 
the kinks out of my neck. 

My first motion picture camera was "ad libbed" out of spare parts 
and a drygoods box swiped from an alley off Hollywood Boulevard. 
It was hand-cranked, the camera. Even then I felt the urge to grow, 
to expand I was very ambitious in those days so we bought a used 
motor for a dollar to run the camera. It had once been a second-hand 
motor, but since that time it had seen everything and died. We had 
to hire a technician to make it go. We have been hiring technicians 
ever since. Our business has grown with and by technical achieve- 

32 W. DISNEY [j. s. M. P. E. 

ments. Should this technical progress ever come to a full stop, pre- 
pare the funeral oration for our medium. That is how dependent we 
artists have become on the new tools and refinements which the 
technicians give us. Sound, Technicolor, the multiplane camera, 
Fantasound, these and a host of other less spectacular contributions 
have been added to the artist's tools, and have made possible the pic- 
tures which are the milestones in our progress. 

That first movie camera now stands in all its ad lib splendor in a 
Los Angeles Museum. Our new multiplane cameras are two stories 
high and operate by remote control. But, on the whole, the basic 
tools and technics of my craft had been worked out before I learned 
the rudiments of animation out of a book in Kansas City. 

There had been animated cartoons long before motion pictures. 
The Stone Age artist came pretty close to animation when he drew 
several sets of legs on his animals, each set showing a different stage 
of a single movement. A Frenchman named Plateau was the first to 
make a cartoon move. In 1831, he invented the phenakistoscope, a 
device of moving disks and peepholes. The successive stages of an 
action were drawn on one disk. When the disk was spun, the illusion 
of motion resulted. Many similar devices were invented to make pic- 
tures move. The first animated cartoon on motion picture film was 
made by J. Stuart Blackton in 1906. It showed a fellow blowing 
smoke in the face of his girl friend. A bit corny, but not bad! Snow 
White and the Seven Dwarfs was not the first feature-length cartoon 
by twenty years, while the first cartoon mechanically colored dates 
back to 1919. The greatest single contribution of the pioneers came 
from Earl Hurd who invented (1915) the idea of tracing the moving 
parts of a cartoon on celluloids superimposed over opaque back- 
grounds. This great labor-saving device is still the foundation of our 
modern method. 

The miracle of seeing drawings move was enough to enthrall the 
early motion picture audiences. Then, as the edge of the miracle 
wore off, interest in cartoons was revived by numerous series of car- 
toons built around the antics of stock characters. Some of these 
series were very popular. Whether or not these pre-Mickey cartoon- 
ists ever sat back and thought about the possibilities in the medium, 
I don't know. I was ambitious and wanted to make better pictures, 
but the length of my foresight is measured by this admission : Even 
as late as 1930, my ambition was to be able to make cartoons as good 
as the Aesop's Fables series. 

Jan., 1941] GROWING PAINS 33 

I was knocking out a series called Oswald the Lucky Rabbit for 
Universal at the time sound exploded like a bomb under silent pic- 
tures. The series was going over. We had built up a little organiza- 
tion. Roy and I each had our own homes and a "flivver." We had 
money in the bank and security. But we didn't like the looks of the 
future. The cartoon business didn't seem to be going anywhere ex- 
cept in circles. The pictures were kicked out in a hurry and made to 
a price. Money was the only object. Cartoons had become the 
shabby Cinderella of the picture industry. They were thrown in for 
nothing as a bonus to exhibitors buying features. I resented that. 
Some of the possibilities in the cartoon medium had begun to dawn 
on me. And at the same time we saw that the medium was dying. 
You could feel rigor mortis setting in. I could feel it in myself. Yet 
with more money and time, I felt we could make better pictures and 
shake ourselves out of the rut. When our distributor, Universal, 
wouldn't give us the money, we quit. Most of our staff went over to 
Universal. That hurt! But I had made my Declaration of Inde- 
pendence and traded security for self-respect. An artist who wouldn't 
is a dead mackerel. Thereafter, we were to make pictures for quality 
and not for price. The public has been willing to pay for this 

Out on my own again, I looked for a new character and hit on 
Mickey Mouse. The first two Mickey Mouse pictures were silent. 
We couldn't peddle them. It occurred to me that in a world gone 
sound-mad, since the release of Al Jolson's The Jazz Singer, a cartoon 
with action synchronized to sound would be something of a sensa- 
tion. My third Mickey, Steamboat Willie, was planned with this in 
mind. By some miracle we managed to figure out the basic method 
for synchronizing sound and action that we still use. When the pic- 
ture was half finished, we had a showing with sound. A couple of my 
boys could read music and one of them could play a mouth organ. 
We put them in a room where they could not see the screen and ar- 
ranged to pipe their sound into the room where our wives and friends 
were going to see the picture. The boys worked from a music and 
sound-effects score. After several false starts, sound and action got 
off with the gun. The mouth-organist played the tune, the rest of us 
in the sound department bammed tin pans and blew slide whistles 
on the beat. The synchronism was pretty close. The effect on our 
little audience was nothing less than electric. They responded al- 
most instinctively to this union of sound and motion. I thought they 

34 W. DISNEY [j. s. M. P. E. 

were kidding me. So they put me in the audience and ran the action 
again. It was terrible, but it was wonderful! And it was something 

I took Steamboat Willie to New York and started a dreary hunt for 
a sound company which was not too busy or too expensive to record 
the sound for me. I finally made a deal with Cinephone. Theirs was 
a pretty punk sound system until Bill Garity redesigned it later on. 
But in spite of that, Steamboat Willie was an instant hit. It played 
the Colony, then moved to Roxy's. Mickey was a big shot over 
night. Lush offers poured in from Hollywood, but Cinephone had 
us nailed to a contract. A year later, in a joint deal with Columbia, 
we bought up the contract. Cinephone had given me a bigger pic- 
ture budget than had Universal, and Columbia had upped the figure 
considerably again. But soon the increasing quality on which we 
were building our business demanded bigger and bigger advances. 
Columbia couldn't take it, so in 1931, we made a deal with United 
Artists to distribute our cartoons. 

This new deal, for all practical purposes, gave us financial inde- 
pendence. Since then, we alone have determined how much our 
pictures will cost. Not that the industry hasn't had a great deal to 
say about our picture costs, in one sense. Time and again, it has been 
said that we were crazy and would go broke. Mack Sennett claimed 
that we put live-action shorts out of business because they could not 
afford to spend the money to compete with us. The fact was the 
reverse. Live-action shorts could not afford not to spend more money 
if it would improve their quality. By 1931, production costs had 
risen from $5400 to $13,500 per cartoon. This was an unheard of and 
outrageous thing, it seemed. And a year later, when we turned down 
Carl Laemmle's offer to advance us $15,000 on each picture, he told 
me quite frankly that I was headed for bankruptcy. This was not 
short-sighted on his part. He had no way of seeing what we saw in 
the future of the medium. 

As Mickey Mouse became a universal favorite and the money 
rolled in, we had been able to afford the time and money to analyze 
our craft. I think it is astounding that we were the first group of 
animators, so far as I can learn, who ever had the chance to study 
their own work and correct its errors before it reached the screen. 
In our little studio on Hyperion Street, every foot of rough animation 
was projected on the screen for analysis, and every foot was drawn 
and redrawn until we could say, "This is the best that we can do." 

Jan., 1941] GROWING PAINS 35 

We had become perfectionists, and as nothing is ever perfect in this 
business, we were continually dissatisfied. 

In fact, our studio had become more like a school than a business. 
As a result, our characters were beginning to act and behave in 
general like real persons. Because of this we could begin to put real 
feeling and charm in our characterization. After all, you can't ex- 
pect charm from animated sticks, and that's about what Mickey 
Mouse was in his first pictures. We were growing as craftsmen, 
through study, self-criticism, and experiment. In this way, the in- 
herent possibilities in our medium were dug into and brought to light. 
Each year we could handle a wider range of story material, attempt 
things we would not have dreamed of tackling the year before. I 
claim that this is not genius or even remarkable. It is the way men 
build a sound business of any kind sweat, intelligence, and love of 
the job. Viewed in this light of steady /intelligent growth, there is 
nothing remarkable about the Three Little Pigs or even Fantasia 
they become inevitable. 

The Silly Symphony series was launched in 1929. In Mickey 
Mouse cartoons, we kidded the modern scene. The material was 
limited. We wanted a series which would let us go in for more of the 
fantastic and fabulous and lyric stuff. The Silly Symphony didn't 
give Mickey much competition until we added Technicolor in 1932. 
We thought that color would be worth the heavy extra cost. Color 
was part of life. A black-and-white print looked as drab alongside 
Flowers and Trees, as a gray day alongside a rainbow. We could do 
things with color! We could do many things with color that no other 
medium could do. 

I remember Roy coming into the office about this time with a bunch 
of figures in one hand and eyes full of patient resignation. "How 
come," began Roy, "How come that last year with thirty men we 
made thirty pictures, and this year with over a hundred and fifty, 
you get out only eighteen?" I can't answer that type of question, 
but the surest way to take Roy's mind off past and present troubles 
is to tell him that we need a lot more money in the immediate future. 
Roy has the greatest confidence in me, in our medium and in our 
future, but he is a business man and doesn't like to live dangerously 
twelve months out of the year. In this instance, three little pigs and 
a big, bad wolf were soon to bring him days of peace not many days, 
but a few. 

The Three Little Pigs was released in 1933. It caused no excite- 

36 W. DISNEY [j. s. M. P. E. 

ment at its Radio City premiere. In fact, many critics preferred 
Noah's Ark which was released about the same time. I was told that 
some exhibitors and even United Artists considered The Pigs a 
"cheater" because it had only four characters in it. The picture 
bounced back to fame from the neighborhood theaters. Possibly 
more people have seen The Pigs than any other picture, long or short, 
ever made. So you get an insight into the short-subject business when 
I tell you that The Pigs grossed only $125,000 its first year. Snow 
White grossed over seven million. That's the difference between 
shorts and features from the profit angle. The low rentals for short 
subjects has been a chronic headache for us. Our only solution has 
been to build our prestige through quality to the point where public 
demand forced the exhibitor to pay more for our product. Theaters 
paying two or three thousand a week for a feature may pay us only a 
hundred or a hundred and fifty dollars for a short. Gentlemen, I ask 
for justice. 

Whatever the reason for The Pigs' astonishing popularity, it was 
an important landmark in our growth. It nailed our prestige way up 
there. It brought us honors and recognition all over the world and 
turned the attention of young artists and distinguished older artists 
to our medium as a worthwhile outlet for their talents. We needed 
these men for future growth, and they came from all over the country 
to join our staff and be trained in our ways. 

The success of the Three Pigs was felt throughout our entire busi- 
ness. The income from all our pictures and from merchandising 
royalties took a sharp up-swing. The magazine Fortune declared that 
our net profit for 1934 was $600,000 and I'll take their word for it. 
That's chickenfeed in Hollywood, but we are strictly small fry. We 
poured the money back into the business in a long-range expansion 
program pointing at feature-length production and the protection of 
our new prestige through constantly increasing quality. The Mickeys 
went Technicolor. We enlarged our training school and began a na- 
tion-wide advertising campaign for young artists. The production 
costs on our Symphonies shot skyward until some of the little pictures 
approached the ridiculous figure of $100,000. But the quality was 
there, and by 1935 even the Three Little Pigs looked dated and a bit 
shabby in comparison with the newer Symphonies. Our staff at this 
time numbered around three hundred. A greater degree of specializa- 
tion was setting in. The plant was becoming more like a Ford fac- 
tory, but our moving parts were more complex than cogs human 

Jan., 19411 GROWING PAINS 37 

beings, each with his own temperament and values who must be 
weighed and fitted into his proper place. I think I was learning a 
great deal about handling men ; or perhaps the men were learning how 
to handle me. But let me tell you this young artists are just as 
reasonable and easy to handle as anybody else. Our temperament 
goes into our job. 

We had our technic well in hand. We had learned how to use our 
tools and how to make our characters act convincingly. We had 
learned a lot about staging and camera angles. We knew something 
about timing and tempo. But a good story idea, in our business, is an 
imponderable thing. It seems to be largely made up of luck and 
inspiration. It must be exceedingly simple to be told in seven or eight 
hundred feet. It must, above all, have that elusive quality 
called charm. It must be unsophisticated, universal in its appeal and 
a lot of other things you can't nail down in words but can only feel 
intuitively. The Three Little Pigs, The Flying Mouse, and The Grass- 
hopper and The Ants, were examples of good stories. I used to feel 
at times that there wasn't another good story idea left in the world 
which could be told in eight hundred feet. The length limitation of 
the Symphony became more and more galling. We were batting 
story ideas around for months and sometimes years trying to get the 
certain twist, the lacking element, or whatever the idea needed to 
make it a good story. Our files were filled with abandoned stories 
on which we had spent thousands. It was inevitable that we should 
go into feature-length pictures if only for the unlimited new story 
material this field held for us. 

I thought we could make Snow White for around $250,000. At 
least that's what I told Roy. The figure didn't make sense because 
we were spending about that much on every three Symphonies or 
2500 feet of picture. Roy was very brave and manly until the costs 
passed a million. He wasn't used to figures of over a hundred thou- 
sand at that time. The extra cipher threw him. When costs passed 
the one and one-half million mark, Roy didn't even bat an eye. He 
couldn't; he was paralyzed. And I didn't feel very full-blooded, 
either. We considered changing the name of the picture from Snow 
White to Frankenstein. I believe that the final figure, including prints, 
exploitation, etc., was around two million. We sort of half-way ex- 
plained this to everybody by charging a million of it off to research 
and development. You know, building toward the future. And this 
was true, although we hadn't exactly planned it to be that way. 

38 W. DISNEY [j. s. M. P. E. 

Webster sums up the spirit of the Snow White enterprise in his defini- 
tion of adventure at the beginning of this article "risk, jeopardy; 
encountering of hazardous enterprise; a daring feat; a bold under- 
taking in which the issue hangs on unforeseen events, etc." 

As a matter of fact, we were practically forced into the feature 
field. We not only had to have its new story material, but also we 
had to have feature profits to justify our continuing expansion, and 
we sensed that we had gone about as far as we could in the short- 
subject field without getting ourselves in a rut. We needed this new 
adventure, this "kick in the pants," to jar loose some new enthusiasm 
and inspiration. 

Research and preliminary work in a small way had begun on Snow 
White as early as 1934. I picked that story because it was well known 
and I knew we could do something with seven "screwy" dwarfs. Be- 
yond that, we didn't know exactly where we were going, but we were 
on our way. The picture was released at the turn of the year, 1937- 
38. At the end of its first year, Snow White and the Seven Dwarfs 
was reported to be the biggest money-maker of all times. It at least 
settled the question as to whether or not an audience could or would 
sit through an hour and thirty minutes of animated pictures. Most 
of the bets were that an audience would go blind. As a matter of fact, 
that question had been settled as early as 1935, when European 
audiences lined up in long queues to see a two-hour bill of our shorts. 
This bill ran for seventeen weeks in Stockholm, and similar all-cartoon 
bills have been quite successful in this country. 

At the time of Snow White's release, our staff had grown to about 
six hundred. Having committed ourselves to a program of both fea- 
tures and shorts, it became necessary again to expand drastically. 
An additional eight hundred people were added to our payrolls in the 
next two years. For more studio space, we were forced to lease a row 
of apartment houses adjoining the studio, and other temporary build- 
ings were erected on the lot. We needed a new studio and in a hurry. 
Not only did we need more space and more buildings, but the in- 
creasing emphasis on the technical side of our craft demanded the 
most modern and specially designed type of buildings and equipment. 
The new plant was started in 1939 on fifty-one acres near the Los 
Angeles River in Burbank. We moved in around the first of 1940. 

The two years between Snow White and Pinocchio were years of 
confusion, swift expansion, reorganization. Hundreds of young people 
were being trained and fitted into a machine for the manufacture of 

Jan., 1941] GROWING PAINS 39 

entertainment which had become bewilderingly complex. And this 
machine had been redesigned almost overnight from one for turning 
out short subjects into one aimed mainly at increased feature produc- 

Produced under such conditions and forced to bear its share of this 
tremendously increased overhead during a two-year period, Pinocchio 
cost something over three million dollars. Suddenly, the world war 
wiped out half our markets. Pinocchio is yet to return its original 
investment. It has been called a flop. Actually it was the second 
biggest box-office attraction of the year. Gone with the Wind was 
first. Pinocchio might have lacked Snow White's heart appeal, but 
technically and artistically it was superior. It indicated that we had 
grown considerably as craftsmen as well as having grown big in plant 
and numbers, a growth that is only important in proportion to the 
quality it adds to our product in the long run. 

The large profits from Snow White, short subjects, and the mount- 
ing royalties from our merchandising enterprises, had all gone back 
into the business to pay for the new studio and expansion program. 
Our payroll had risen to around three million a year. The war had 
cut our potential picture profits in half. The crisis was on. Another 
one. It was brought on by what might reasonably be called reckless 
expenditures. Yet, looking at it our way, it is these expenditures 
that have put us in shape for the storm. Instead of the one feature- 
length picture every two years which seemed the limit of our capacity 
two years ago, we are now reorganized and equipped to release nine 
features in the next two years, each at a fraction of Pinocchio' s cost. 

The first of these nine features, Fantasia, has been released. We 
have never been so enthusiastic about a picture. Every picture is an 
adventure, but Fantasia has certainly been our most exciting one. 
We take great music and visualize the stories and pictures which the 
music suggests to our imaginations. It is like seeing a concert. Leo- 
pold Stokowski and the Philadelphia Orchestra recorded the music, 
using a new system of sound recording and three-dimensional re- 
production called Fantasound. It is our intention to make a new 
version of Fantasia every year. It's pattern is very flexible and fun 
to work with not really a concert, not vaudeville or a revue, but a 
grand mixture of comedy, fantasy, ballet, drama, impressionism, 
color, sound, and epic fury. 

Mickey Mouse and Disney in the same boat with Bach, Beethoven, 
Stravinsky, and Stokowski! Well, where do we go from there? I 


haven't the faintest idea. I have never had the faintest idea where 
this business would drag me from one year to the next. It's at the 
controls, not me ! But, as I said before, as long as we keep on growing 
the future will keep opening up. More than any other picture, 
Fantasia shows how much the medium has grown. No doubt, some 
unimaginative critics will predict that in Fantasia the animated 
medium and my artists have reached their ultimate. The truth is 
quite to the contrary. Fantasia merely makes our other pictures 
look immature, and suggests for the first time what the future of 
this medium may well turn out to be. What I see way off there is too 
nebulous to describe. But it looks big and glittering. That's what 
I like about this business, the certainty that there is always something 
bigger and more exciting just around the bend ; and the uncertainty 
of everything else. 

Over at our entertainment factory we are training hundreds of 
brilliant youngsters to carry on the job far beyond where we old 
timers must leave off. They will train other youngsters. There is no 
knowing how far steady growth will take the medium, if only the 
technicians continue to give us new and better tools. For the near 
future, I can practically promise a third-dimensional effect in our 
moving characters. Fully exploited, Fantasound should prove a 
startling novelty. The full inspiration and vitality in our animators' 
pencil drawings will be brought to the screen in a few years through 
the elimination of the inking process. Then, too, our medium is 
peculiarly adaptable to television, and I understand that it is already 
possible to televise in color. Quite an exciting prospect, I should 
say! And, since Fantasia, we have good reason to hope that great 
composers will write directly for our medium just as they now write 
for ballet and opera. This is the promise of the next few years. Be- 
yond that is the future which we can not see, today. We, the last of 
the pioneers and the first of the moderns, will not live to see this 
future realized. We are happy in the job of building its foundations. 


To pioneer has always been with me an obsession. Perhaps the 
yearning to explore new fields was an inheritance from colonial an- 
cestors. Vanished geographical frontiers still left far vaster regions 
in science and technology to explore. When early wireless began to 
be a bit crowded the radio telephone field, then scarcely a dream even 
among communication engineers, beckoned me irresistibly. This 
primitive beginning of the radio broadcast, in 1908, logically necessi- 
tated the development of the electronic amplifier from the audio n 
detector tube (Fig. 1) ; and thus again I managed to escape the crowd. 
And when, in 1912, this amplifier proved to be also an oscillator, a 
boundless ocean disclosing alluring archipelagoes of practical applica- 
tion was opened to scientific research. 

It then became apparent that many a forgotten dream of other 
early inventors might finally be brought to realization. Among such 
in the sea of television were the scanning disk of Nipkow, the cathode- 
beam picture of Rossing, Cambell-Swinton's invention of the cath- 
ode-scanning beam (recently perfected in Zworykin's " Iconoscope") 
all brilliant conceptions which must needs remain only blueprints 
and Letters Patent, for the simple lack of an inertialess amplifier of a 
hundred million magnifying power. 

Similarly in acoustics the primitive but all-embracing patent of 
Fritts, breaking all records, embalmed for thirty-six years in the 
Patent Office; and that of Elias Ries who in 1913, before the photo- 
electric cell, or the amplifier which could make it useful, described the 
remaining essentials of photographic sound-on-film recording and 

In 1919, happily unconscious of these then buried documents, I de- 
cided the time had at last arrived when sound photography should 
definitely give a voice to the picture film. 

*Requested and recommended for publication by the Historical Committee. 
""Hollywood, Calif. 




tf. S. M. p. E. 

I essayed at first three methods of sound recording, the speaking 
flame, the tiny incandescent filament, and the glow- tube. The latter 
soon showed itself to offer the only hope of practical success (Fig. 2). 

My first demonstrated actual combination of sound-on-film and 
talking picture was at my old High Bridge, New York, laboratory in 
the spring of 1921, shortly before I removed to Berlin. My then 
assistant, William Garity, still cherishes a few film samples of him- 
self holding the hand microphone, while I served as cameraman. 

This early work, when apparently only we two (and he, somewhat 
skeptically) believed there was a commercial future for the talking 
picture, evidently sank deep within his soul; for today Garity is chief 

FIG. 1. The De Forest 
audion original type. 

factotum for Walt Disney; possibly because that primitive recording 
was chiefly suggestive to him of the squeaks of Mickey Mouse. 

It was in the spring of that year, 1921, that my difficulties in de- 
veloping properly a sadly underexposed sound record and overex- 
posed picture on the same film suggested the use of two separate, syn- 
chronized negatives, one for the picture, one for the sound, each given 
its proper development, and each printed successively on a common 

My patent application covering this basic principle was finally re- 
jected after a bitterly contested interference proceeding with that by 
the Tri-Ergon inventors. The destiny of this latter patent in our 
Supreme Court is now recent history, familiar to all. I still maintain, 
however, that here resided a genuine invention, once a total novelty, 
and now of tremendous practical value. 

Shortly after my return to America a year later, in 1922, and my in- 
stallation in a genuine motion picture studio, that ancient remodelled 
brewery of Tec-Art on East 48th Street, I was visited by Theodore 


Case of Auburn, N. Y. He watched my work and shortly thereafter 
summoned me to his Laboratory to show me a gassy Western Electric 
amplifier bulb whose "blue haze" was fluttering in accord with tele- 
phone currents from his microphone. Forthwith I sketched out the 
first oxide-coated cathode glow-tube, which he and E. I. Sponable, 
his gifted assistant, constructed and named the "AEO light" (Dec., 
1922) ; whereupon I proceeded to scrap my radio-frequency recording 
oscillator and metal-cathode glow-tubes in favor of this low-voltage 

FIG. 2. Lee de Forest with two early types of 
glow- tubes for sound-on-film recording (1925). 

direct-current source. I also discarded my Kuntz photoelectric cells, 
difficult to obtain with uniform quality, in favor of the Case far more 
sensitive "Thalafide" (resistance) cell, enthusiastically regardless of 
the fact that the latter cut off quite effectively below 3000 cycles. 

After this experience I returned to the use of the photoelectric cell 
with resultant gain in quality, and to the use of the metal cathode re- 
cording light, but designed to operate on low voltages, thus obviating 
the use of the radio-frequency oscillator of my first system. I shall 
refer to this feature hereafter. 

44 L. DE FOREST [J. S. M. P. E. 

But now the more directly commercial requirements, following 
upon my introduction of the "Phonofilm" to Broadway audiences 



Operated in conjunction with THE RIAI.TO, Times Square 

KIKE NOTlCK-I.ook around NOW and choose the nearest exit to 
your seat. In awe of tire walk (not run) to THAT EXIT. Do not 
try to beat your neighbor to the street. 

THOMAS J. URENNAN. Fire Commissioner. 

f K (J It A M 



EVENINGS Orchestra 83 cent* 

Admission 77 cents 
War Tax 8 cents 

Balcony SO cents 

Admission 43 cents 
War Tax 3 cents 

Evening prices 
prevail at Satur- 

MATINEES Orchestra 50 cents 

Admission 43 cents 
War Tax 5 cents 

day, Sunday and 

Balcony 80 cents 

Admission fl cents 
Wnr Tax a r+ntm 

Holiday Matinees 

LOOES Always .. cent,) # } g 

The Rlvoli opens at noon dally, performances begin at 12 noon. 2:13*, 
4:13. 7:30* and 9:30 p. in. .' Sunday first performance starts at I. 
Asterisks indicate full presentation with orchestra, soloists and scenic 
effect*. J. Van (left Cooper nnd Frank Stewart Adams at the organ. 

PROGRAM Witk of April 15th. 1!>'j:i 

FRKDHKICK STAHI.nF.RO, Conductor WILl.IF. STAIIL. Asst. Conductor 

I. Ovi'.irrritK 


Jacques OtTcnliach 
STMIIHHU; and WII.IIK Si A ill. condildtin^r 

Although Orpheus is not mentioned in the Homeric poems. the 
story of his attempting to bring back hU dead wife, Kiirydicc. 
from the lower regions was a favorite of the (irceks and hits 
Itecn ever since. It w is a subject of the first grand opera ever 
written, by I'erl in Iflnit. and is the theme of Murk's most famou* 
opera, produced in 1702. Offenbach's hiirlcst|iie of the old myth 
was produced in Paris in I>OM. and its jolly time made it a great 
success. The overture is one of the finest specimens of light nuisir. 

TIIK Pimxoni.M 

'/'/* riiniitijilin in tlif l<ite*t invent inn of />r. l.ff ilr l-'nrrl 
ami in n hnni ittriiir foru'nrtl in the <l<rveli>innfnt of the ln)k- 
imi ftlrliiren. l-"i>r the jir*t time it Im* heen matle fmnnihlg 
In rtcnnl the ftirfure titnl the mire or iitn.*!r int the *<i>n* 
film, flnin iiinuriiiif perfect tynrkriHtiZHtiun. Thin ix tjie firnt 

FIG. 3. Section of Rivoli Theater (N. Y.) program for 
the week of April, 15, 1923. announcing de Forest phono- 
film demonstration. 

under the far-visioned sponsorship of Dr. Riesenfeld, at the Rivoli 
and Rialto Theaters on April 15, 1923 (Figs. 3 and 4) resulted in 
filing early patent applications in 1923-25 on such extremely practical 


inventions as these : the use of two or more picture cameras at differ- 
ent angles and focal distances, all synchronized to a common sound- 
recording camera; the blacking out by printing, of the otherwise 
noisy pauses in the positive sound-track (the basic patent in all proc- 
esses of "noiseless recording"); the method of dubbing sound re- 
corded in synchronism with a projected picture (Fig. 5). This was 
practiced first in 1924 for The Covered Wagon, as exhibited in the Rivoli 
Theater during portions of the "Supper Shows" when Riesenfeld's 
Orchestra was not playing. We did the same in 1925 for his splendid 

FIG. 4. Rivoli (New York) Theater projection room, Phonofilm equipped 


score of Siegfried, actually recorded in the Century Theater while the 
orchestra was playing to the projected picture. 

Monologue numbers by Eddie Cantor, George Jessel, DeWolf 
Hopper and Chic Sale; dialogs between Gloria Swanson and Thomas 
Meighan, and Weber and Fields; Fokina's Swan Dance; playlets 
with Raymond Hitchcock ; orchestra recordings by Ben Bernie, Paul 
Specht, Otto Wolf Kahn, and similar entertainment made up our 
repertoire during this early period of 1923-27 (Fig. 6). 

It may not be remembered that Technicolor was first wedded to 
sound in the spring of 1925, when Balieff's entire Chauve Souris was 
thus recorded, using a sound-camera synchronized to the color camera, 

46 L. DE FOREST [j. s. M. p. E. 

whose noise was quite successfully suppressed beneath a quilt blimp 
within an "ice-box" booth with walls eight inches thick. The sound- 
track was then printed on the green positive, which dyed surface, al- 
though serving better than the red, was found quite unsuitable. 

Nevertheless certain numbers of this production were exhibited 
during 1924-25 to enthusiastic audiences in the London Tivoli, 
and in Japan and Australia. 

About this same time I adopted as standard back-screen equipment 
large vertical exponential horns of wood with a cone-speaker at the 

FIG. 5. De Forest sound-on-separate film cameras (1924) . 

base, and trumpets with Western Electric dynamic receivers located 
in the bell of the horn. Theater screens were yet unperforated. 

Another very practical patent taken out during this era covers the 
camera "blimp," in an acoustically treated studio. 

"Phonofilm" reproducing apparatus installed in thirty-four theaters 
scattered throughout the East in 1924-25 led to the joint invention by 
Louis Reynolds and the author of the now well known "tone-control," 
whereby the operator, or a monitor in the auditorium itself, was en- 
abled to mix the relative values of high and low frequencies to suit 
best the acoustic characteristics of the theater, or as the audience 
grew or diminished. 


My early theatrical experiences were replete with humorous inci- 
dents, interspersed with discouragement and heart-failure. For ex- 
ample, the only occasion on record when one of my first operators, 
Billy Brinkman, failed to forget to turn on the amplifier switch before 
he was applauded or kicked into that action was when he went home 
after the show one night and forgot to switch the amplifier off! The 
following matinee went off without a hitch. Unquestionably I 
shortened my life by several years by dashing up long flights of gallery 
stairs, two at a stride, to endeavor to start or improve the sound re- 

FIG. 6. Entrance to Capitol Theatre, New York (1927). 

The spring of 1924 witnessed the first talking Newsreel, when an 
improvised sound-truck journeyed to Washington to record a pre- 
campaign talking-picture of President Coolidge on the White House 
lawn. That same summer saw also the Progressive Candidate, 
Senator La Follette, and the Democratic, John W. Davis, vieing for 
motion picture theater audience popularity. In 1925 Al Smith and 
"T. R., Jr.," each visited my studio to tell the recording camera why 
he should be elected Governor of New York. 

The early public acceptance of this type of News Weekly readily 
convinced Manny Cohen, of Pathe News, that it possessed an assured 
future. Only our exceedingly modest royalty demands prevented 

48 L. DE FOREST [j. s. M. p. E. 

him from thereby saving millions in later royalties for that organiza- 

Today, when the entire motion picture world has been for several 
years almost 100 per cent "talkie," it seems to me incredible that 
only a little more than a decade ago not one of the "Big Guns" in 
the industry believed that the talking picture had any place in the 
theater. Such "best minds" as Adolf Zukor, Sidney Kent, Gold- 
stein, et al., turned a deaf ear to all arguments by myself, Hugo Riesen- 
feld, and Harold Franklin. When I found that William Fox was a 
fellow passenger on the Berengaria as I returned with my demon- 
stration equipment from the Berlin Laboratory, he refused to meet 
me even to discuss the subject. And when in 1924 he returned to 
New York and learned that Phonofilm was actually installed in some 
six of the Fox houses he peremptorily ordered them all taken out, 
without even deigning to witness a demonstration. 

Yet a year or so later, when the far-sighted Courtland Smith had, 
almost surreptitiously, installed the Case equipment in the Tenth 
Avenue studio, Fox, then very likely aroused by reports of how 
"Vitaphone would astonish the world," lost no more time in tying 
up with the invention and launched a program which brought into 
being an eight million dollar "Movietone City." 

Even Sam Katz, astute purveyor of the newest and most daringly 
original acts and stunts for entertaining a blase public, turned in 1926 
the glazing eye and the clammy hand to my lieutenants who sought a 
limited- term contract to road-show this "short-lived novelty" of the 
"Talking Picture." 

Unquestionably it was the absolutely unique prescience and cour- 
age of Sam Warner, and later his brothers, which finally resulted in 
arousing the motion picture industry to the belated realization that 
here at last science and invention had created a new instrumentality, 
one which the mute public had long and patiently awaited; and 
which, once launched on the sea of public acceptance, was destined to 
sweep over those antiquated studios and half-empty theaters with a 
tidal wave of irresistible momentum, ruthlessly scrapping their worn- 
out equipment, outdating their time-honored technic, relegating their 
priceless art and high-priced artists to an oft-lamented limbo, at a cost 
in millions which staggered even the intoxicated imagination of Wall 
Street, in the millennium of predepression "rugged individualism." 

But although Vitaphone and phonograph-recording got away first 
in its race with film-recording, the terrific handicaps of its involved 


technic, geometrically increasing as its public acceptance grew, in- 
evitably led to its general abandonment in favor of the numerous 
practical advantages which I regarded at the very commencement of 
my researches as inherent to any photographic sound-on-film process. 

And while on the subject of future developments I venture to 
prophesy an independent volume-control sound-track, shown in an 
early patent, will, I beieve, yet demonstrate its utility in the art. 
The second track offers certain advantages in dubbing, avoiding all 
photographic complications of superimposing a loud record upon a 
weaker, or the reverse. In general, the inherent limitations of the 
emulsion may be entirely eliminated by the use of the double sound- 

Although the author has been for the past five years quite outside 
of the then too crowded talking-picture activities, yet to look back 
upon all this history of invention, this genuine social revolution, this 
Caeserian birth of a national industry, in which it was his fortune 
to pioneer is to me now at least a source of grim satisfaction, im- 
possible to express. 


Summary. Although the new Twentieth Century camera was designed primarily 
to reduce noise, it also embodies many of those conveniences and devices which spell 
speed and aid in cost cutting. 

The camera has been designed and built along new principles and, instead of 
trying to hold the noise in the camera case or the blimp, the noise has been reduced 
at its source. The fast-moving reciprocating parts are as light and as small as pos- 
sible, and when assembled yield uniform acceleration and deceleration, with a re- 
sultant optimum movement of the film and a reduction in noise-making vibration. 
This, when coupled with a patented sound-insulating mount for the film-moving 
mechanism, reduces the noise output to a level substantially equivalent to the noise 
hvel of the best blimped camera available. Other features included in the camera are 
described in the paper. 

Although the Twentieth Century-Fox Film camera has been de- 
signed primarily with the idea of reducing noise so that it may be 
used without a blimp, it also embodies many new devices and ac- 
cessories which make for speed and ease of operation. It may 
properly be called a Cameraman's Camera, since it was built by 
cameramen for cameramen. The authors, together with Robert 
Stevens and the late Charles Miller, who conceived and de- 
signed this camera, at the outset realized that there were two ways 
of reducing noise: one was to insulate against noise transmission by 
using a camera blimp the old way; and the other, or modern way, 
to reduce the noise at its source. They rejected the prevalent idea 
that noise could not be further reduced at its source, and set about 
developing a silenced camera along entirely new lines. The theory 
was that if all moving parts, particularly reciprocating parts, were 
made as small and as light as possible, and were then moved with 
uniform acceleration and deceleration, not only would the minimum 
amount of vibration be set up, thus reducing the noise at its source, 
but an optimum movement of the film would also result. In this 

* Presented at the 1940 Fall Meeting at Hollywood, Calif.; received Septem- 
ber 30, 1940. 

** Twentieth Century-Fox Film Corp., Hollywood, Calif. 


FIG, 1, Two views of the new camera. 

52 D. B. CLARK AND G. LAUBE [j. s. M. P. E. 

camera this theory is utilized to the full extent with amazing results. 
However, it was realized that a certain amount of vibration would 
be set up regardless of construction, and a way was sought to elimi- 
nate even this. The result was a sound-insulating mount for the 
film-moving mechanism, designed to absorb certain of the vibrations 
and to convert other vibrations into inaudible frequencies. This 
mount incorporated in this camera plays a part in reducing the final 
noise output to a level less than the noise level of the best blimped 
camera available today. 

While the design was directed primarily toward noise-reduction, 
there are also embodied in the camera other features dealing with 

FIG. 2. Focusing position. 

improvements in photographic characteristics and conveniences, 
such as a new focusing device, novel lens mounts, follow-focus moni- 
toring finder, improved shutter, new type of film-moving mechanism, 
synchronizing means for background projection shots, sound-proof- 
magazines with film spool lock, hydraulic free-head, lens calibrating 
system, and a slating device, all tending to produce speed and ac- 
curacy in operation. One of the most important of these features 
comprises the focusing means, which is particularly pleasing to di- 
rectors and cameramen. 

Instead of sliding the case laterally, as cameramen are accus- 
tomed to do now for lining up and focusing, the case, with the maga- 
zine, is rotated about the shutter axis. To do this, the case is mounted 
in a yoke and when the magazine on the case is in an upright posi- 

Jan., 1941] 



tion, the camera is in a shooting position (Fig. 2). For focusing, the 
case is tilted 75 degrees about the shutter axis, which removes the 
film-moving mechanism and the film-loop away from the photo- 
graphing aperture and brings a ground-glass and an aligned optical 
system into position behind the aperture. .Under this arrangement, 
focusing is done directly through the photographing lens by means 
of an eyepiece mounted on the rear of the yoke and the optical system 
through the case. The operation is simple, fast, and accurate, and 
the magazine in the inclined position is out of the way, which gives a 
decided advantage for lining up shots. A small handle is mounted 
on the rear of the yoke for rotating the case into and out of shooting 






FIG. 3. Lens mount. 

Next is the lens mount, which is built up of three concentric sleeves : 
an outer sleeve, an inner sleeve, and an intermediate sleeve, and com- 
prises front and rear ball-bearings to support and align the sleeves 
(Fig. 3) . The outer sleeve acts as a support and is adapted to be at- 
tached to the lens turret of a camera; the inner sleeve is the lens 
barrel, and the intermediate sleeve rotates between the other two 
sleeves. The inner sleeve or the lens barrel is splined to the outer 
sleeve to prevent its rotation with the intermediate sleeve and the 
rotary movement of the intermediate sleeve is translated into recti- 
linear movement of the lens barrel by means of the front ball-bearing 
running in helical grooves on the inner and intermediate sleeves. 
The amount of lens movement therefore depends upon the pitch of 
the helical grooves, and the pitch is calculated and constructed to 
yield the proper rectilinear focusing movement for lenses of various 
focal lengths with the same amount of turning of the intermediate 

54 D. B. CLARK AND G. LAUBE [J. S. M. P. E. 

sleeve in all cases. The result is that the focusing segment may be 
standardized for lenses of all focal lengths, since the same amount 
of turning movement is required for focusing all lenses. This move- 
ment, being the same for lenses of all focal lengths, is utilized for 
focusing and training a follow-focus monitoring finder (Fig. 4). 

This finder is operated from the side of the camera in the conven- 
tional manner and is hinged so that it may be swung into an out-of- 
the-way position for opening the camera case. In order to avoid the 
usual parallax, due to the separation of the finder optical system from 


FIG. 4. View showing monitoring finders. 

the camera lens, the finder is equipped with a means for moving the 
finder lens in a diagonal path, so that as the lens is focused its optical 
center is also offset to compensate for the parallax between the 
monitoring system and the camera system. Both the camera lens and 
the monitoring finder are controlled through the conventional focus- 
ing arm on the camera, which not only properly focuses the camera 
lens through the regular focusing link but also focuses the monitoring 
finder lens, and at the same time trains the monitoring lens upon the 
same field as the camera lens by means of a cam and rocker-arm ar- 
rangement, which moves the movable element of the finder in a 
diagonal path. For lenses of different focal lengths, differently 

Jan., 1941] 



shaped cams are used, some of which are seen in Fig. 4. This insures 
that the cameraman will have the same boundaries of the field in the 
monitor as are seen by the camera, regardless of field distance or the 
focal length of the lens. Provision is made for optionally and rapidly 
throwing the monitoring lens back to its original position on infinity 
for a quick reference cut back to the background field of the camera 

In keeping with the noise-reduction purposes of the camera, care- 
ful attention has been given to the shutter. In order to avoid focal- 
plane disturbances due to vibrating air columns, the shutter has been 
formed of very thin material and has been streamlined wherever 
possible. The shutter opening has been made variable and can be 
used from 45 to 200 degrees and 
still fall within the dwell period 
of the film movement. In de- 
signing the shutter and the aper- 
ture, it was found that the posi- 
tion of the aperture with respect 
to the axis of the shutter was an 
important factor. By enlarging 
the shutter and placing the aper- 
ture as far from the shutter axis 
as possible, the opening and clos- 
ing time of the shutter was con- 
siderably reduced. In Fig. 5 the 
location of the Twentieth Cen- 
tury aperture is shown. Due to the wide shutter, its opening and 
closing time is considerably less than half that of the average motion 
picture camera. This feature provides a longer dwell period for the 
film, which enters largely into the photographic improvements of the 
camera and requires a faster type of film-moving mechanism. 

The film-moving mechanism (Fig. 6) embodies some new princi- 
ples and new design. The chief objective in designing the movement 
was, first, to provide the fast film movement required, and at the 
same time hold the movement of the reciprocating parts to a uni- 
formly accelerating and decelerating motion. Other considerations 
were to make the moving parts, particularly the reciprocating parts, 
as light as possible and to provide a construction wherein the bearing 
surfaces would be rounded surfaces of considerable extent with no 
line contact bearings. To meet these conditions, the pivoted arm 

Fig. 5. Diagram of enlarged shutter, 
with aperture placed farther from the 

56 D. B. CLARK AND G. LAUBE [J. S. M. P. E. 

was adopted with a sliding sleeve worked from an eccentric pin. 
This structure not only provided the proper bearing surfaces but 
lent itself admirably to the requirements of the take-down move- 
ment. The take-down pin itself is a very small cam-actuated member 
slidably mounted in the swinging end of the arm (Fig. 7). By 
properly designing the cam movement for the take-down pin and 
correlating this movement with the eccentric movement controlling 
the pivoted arm, the take-down pin may be made to trace a path as 
shown in Fig. 7. 

FIG. 6. Film-moving mechanism. 

Here it will be noted that at the points where the pin actually 
enters and leaves the sprocket-holes, the movement of the pin is sub- 
stantially perpendicular to the film. These two movements occur 
at the time when the take-down arm is substantially at rest. Since 
the film is at rest there is no focal-plane disturbance of the film by 
the sawing action of the pin in entering or withdrawing from the 

The speed of the take-down pin arm has also been carefully de- 
signed and calculated, and is best shown in Fig. 8. This curve of the 
speed of the take-down pin with respect to time shows that the move- 
ment approaches a sinusoidal curve, which is the ideal condition. 

Jan., 1941] 



The curve has a smooth contour and displays no irregularities which 
would denote sudden changes in speed, indicating disturbances that 
would set up audible vibrations. This is brought about by the slid- 
ing sleeve and eccentric pin, in combination with the pivoted arm, 






FIG. 7. Schematic arrangement of 

which produces the uniformly accelerated and decelerated move- 
ment of the take-down pin that not only reduces shock and vibration 
but yields the optimum movement to the film. It might also be 
pointed out that the highest speed of the Twentieth Century camera 
pin movement is under seventy- three inches a second. 


FIG. 8. Velocities of intermittent movements at take- 
down pins. 

In conjunction with the take-down pin are employed two register 
pins which engage the film on opposite sides directly below the aper- 
ture (see Fig. 6). The movement of these register pins has been so 
designed that the pins hold the film in position until the take-down 
pin has engaged the film, whereupon they release and pass control 
of the film to the take-down pin. This correlation of pins has the film 
under control at all times and insures that there can be no movement 

58 D. B. CLARK AND G. LAUBE [J. S. M. P. E. 

of the film after the shutter once starts opening. The register pins 
are operated from a scroll-cam properly designed and correlated 
with the eccentric and the cam that operates the take-down pin to 
insure this movement. Attention is called to the convenient loading 
pins and the loading hole marker. Convenient means is also pro- 
vided for holding the register pins out of engagement for loading pur- 

When shooting background projection shots, the cameraman is 
often worried over the synchronism of the shutters, and whether he 
is getting the full amount of light possible. This camera provides 


FIG. 9. Shooting position. 

a very simple means for checking the position of the camera shutter, 
with respect to the projector shutter by a little synchronizing aper- 
ture (Fig. 9) which is 180 degrees removed from the photographing 
aperture and is positioned within the radius of the shutter. There- 
fore, when the cameraman looks through this aperture and sees 
light on the background projection screen, he knows that the camera 
shutter is substantially 180 degrees out of phase with the projector 
shutter. It is then merely a matter of adjusting the phase relation 
of the shutters until this aperture goes dark. The darkest point 
would indicate the point where the shutters are in correct phase rela- 

Jan., 1941] 



In conjunction with the synchronizing aperture, the camera car- 
ries means for rotating the motor field so as to bring the camera 
shutter into the proper phase relation with the projector. For this 
purpose there is a motor field shift ring on the rear of the motor 
housing, which is arranged through a specially designed structure to 
rotate the motor field. To line up a process shot, it is merely a mat- 
ter of standing behind the camera, peering through the synchronizing 
aperture and turning the motor field shift ring on the back of the 
motor until the synchronizing aperture goes dark, whereupon the 

FIG. 10. The magazines are insulated to reduce noise. 

cameraman knows that the two shutters are in synchronous relation. 
A thumb-nut locks the field in position. 

As a means toward the end of reducing noise, the magazines were 
specially treated with interior and exterior insulation (Fig. 10). In 
addition to these precautions, an anti-rattle film spool spindle was 
designed. This spindle clamps the film spool firmly and not only 
prevents the spool from rattling, but by eliminating wobbly move- 
ment of the spool tends to force the film to pile up in smooth reels, 
thus preventing any side scraping. 

Having built a camera that was believed would meet the stringent 
requirements of directors of photography, the inventors and engineers 
set about to design a worthy accessory in the form of a new free-head. 

60 D. B. CLARK AND G. LAUBE [J. S. M. P. E. 

In keeping with the novelty of the camera, they discarded the old 
idea of frictional free-heads and conceived and designed a free-head 
(Fig. 11) wherein no friction is employed for resisting either the tilt- 
ing or the panning movement. Both these movements are made 
against a precisely controlled hydraulic resistance. The hydraulic 
resistance used in these movements provides a silky, smooth drag 
having no jerkiness due to friction suddenly letting loose, or uneven- 
ness due to uneven wearing of the friction plates, no matter how big 
the swing or how fast or slow the movement. There is a distinct dif- 
ference in the feel of this movement over the frictional movement, 

FIG. 11. Close-up of free-head. 

which can not be appreciated until once it has been experienced. 
The tilting control is on the side of the free-head yoke and the pan- 
ning control is directly beneath the free-head in the apex of the tri- 
pod. Both movements are locked by the convenient lever nuts shown 
in the photograph. 

In order further to smooth the path of the cameraman and make 
the practical use of photoelectric meters possible, the old method 
of calibrating the light stops on lenses was discarded and a new and 
revolutionary method inaugurated. This new method is based upon 
the proposition that the only true way of rating the light-transmitting 
capacity of any lens is by measuring the actual amount of effective 
light transmitted through the lens. The old method, whereby lenses 

Jan., 1941] 



were rated with respect to light speed by the ratio of the aperture 
opening to the focal length, did not take into consideration the 
human error in measuring the aperture, the number of elements in 
the lens, the arrangement of the elements, the different compositions 
of the glass, or the light lost by reflection and absorption. All these 
physical characteristics enter into the light-transmission capacity 
of any lens. Consequently, under the //system, lenses of different 
makes and different focal lengths would not transmit the same 
amount of light under the same light stop. This fact became more 
and more evident as lenses of different makes and different focal 
lengths were interchanged in a procession of shots during a day's 
shooting, and the cameraman found it becoming more and more 
difficult to match negative densities in his day's shooting. Under 
the supervision of D. B. Clark, one of the authors, all Twentieth Cen- 

c A ULTU ttMsmw 

FIG. 12. Illustrating method of calibrating lenses. 

tury lenses have been calibrated with respect to light transmission 
and rated according to the actual amount of effective light trans- 
mitted. The ratings of the different apertures are all referred to a 
standard light base, which has been established and maintained for 
lens calibration. The means for doing this is schematically shown 
in Fig. 12, and comprises a source of light which is diffused and main- 
tained at a fixed level and a lens holder arranged to be placed closely 
in front of the source of light. In the lens holder there is another 
screen, preferably a ground-glass, which receives the focal image of 
the lens, and back of this screen is a photoresponsive tube of the 
voltaic type. The output of the tube is connected to an ultrasensi- 
tive meter. The whole system is standardized by measuring the 
light transmitted through a carefully selected reference lens set at 
the conventional //3.2. The meter reading for this setting is used 
as the standard reference level for all lenses and the system is cali- 
brated from this reference lens and checked against this lens during 

62 D. B. CLARK AND G. LAUBE ij. s. M. P. E. 

the calibration of other lenses, thus minimizing errors from fluctua- 
tions in the light-source or the output of the photoelectric cell. 

In calibrating a lens, the lens is mounted in the holder. The dia- 
phragm control is then operated until the meter reads the same level 
as it did on the standard setting. This would be the //3.2 reading 
on this lens, and light stops in both directions away from this point 
may be made and calibrated on the lens by multiples of the current 
shown on the meter. It will be noted that the lens holder may be 
moved with respect to the light-source, the only requirement being 
that the field subtended by the lens does not exceed the uniformly 

FIG. 13. Shooting position of slater. 

lighted field source. The equipment shown in Fig. 12 may be used 
in daylight upon a bench in a laboratory and will yield sufficiently 
accurate results to calibrate a lens system under this new method 
whereby photographic results may be mechanically improved and 
negative densities may be more perfectly matched. The success of this 
system is evidenced by the uniform results obtained by all camera- 
men who have worked with lenses so calibrated. Particular success 
has been had on exteriors where light stops are numerous, and even 
more so when working with coated lenses which have a high light 
transmission characteristic, regardless of the make or design of the 

Probably one of the most useful camera accessories developed in 
recent years is a little semi-automatic slating device (Fig. 13), de- 

Jan., 1941] 



veloped by D. B. Clark and colleagues at Twentieth Century-Fox 
Studio. This little device is a complete unit in itself and carries its 
own optical system, its own illumination, and means for mounting 
the indicia required to slate the film. In the photograph showing the 
camera shooting position, it will be noted that the device is suspended 
on the matte box bracket, and when not in use hangs inconspicu- 



FIG. 14. Slating device. 

ously below the sunshade. From this position it is merely a twist 
of the wrist to throw the slating device upward and around into the 
sunshade a few inches in front of the camera lens before the camera 
starts turning (Fig. 14), and, since the device is arranged to block off 
all foreign light from the camera lens and automatically supply its 
own illumination, the first frame of the slate on the film can be used 
as a cue mark. Furthermore, the slate is exposed while the camera 
is coming up to speed, with no interference of the actors on the set. 


This method saves considerable time as well as the film ordinarily 
wasted in bringing the camera up to speed, and does away with the 
disconcerting practice of a slate boy or camera assistant thrusting a 
slate in the faces of the players after they are all set for action. 

The miniature slate is easily removed and numbers are quickly 
changed with a pick provided, or any pointed instrument (Fig. 15). 
Space is provided for the names of the director, cameraman, and 
sound man, production number, camera number, date, scene, take 

FIG. 15. Removing miniature slater from slating 

number, sound-track, playback, and start. These, of course, can 
be varied to meet requirements. The indicia are placed on a plane 
surface, including the number faces, which provides a flat reflecting 
surface for projecting them through the optical system on to the film. 
Although this camera has been designed primarily to reduce noise, 
nevertheless the authors believe that, together with the accessories, 
it eliminates many inconveniences and annoyances that the camera- 
men have so long encountered, and places in their hands a device 
which offers the maximum in economy and efficiency, as well as 
paving the way for a greater margin of excellency in photographic 


Summary. This paper discusses some of the problems involved in the con- 
sideration of suitable standards now before the National Television Systems Com- 
mittee. Resolution is approached from a standpoint of the number of lines and fields 
within the limits of presently assigned channels. Related problems touched upon are 
flicker frequency vs. illumination, and some of the difficulties which must be guarded 
against with colored images such as rainbow effects and flicker, both of which can be 
minimized by using relatively high frame frequencies. The effects of motion, which 
tend to smear detail, are discussed in relation to frame and field frequency. 

The major limitations of present scanning spot shape and intensity distribution, 
which determine the vertical and horizontal widths of confusion, have been removed 
in the laboratory, introducing the possiblity of markedly improved definition with a 
given number of lines and fields, which must be reckoned with in determining stand- 

The purpose of this paper is to create a deeper appreciation of 
some of the problems in television which need to be settled by mutual 
agreement within the industry before we can expect commercializa- 
tion, which, in turn, will inevitably be followed by an expansion to a 
national service of a medium which has much to offer the American 
public. In the space allotted for this discussion of image resolution 
only the primary factors can receive any attention. 

We will make the broad assumption that, regardless of the number 
of frames, fields, or lines and element areas, items such as the follow- 
ing are roughly equivalent for the various systems discussed : 

(1) Contrast range (roughly equivalent to gamma). 

(2) Visual response and resolution as a function of brightness. 
(5) Overall phase delay linear. 

(4) Amplitude response and band width. 

(5) Shading. 

(6) Screen color. 

(7) Spot shape distortion. 

(8) Thermal noise. 

* Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received October 25, 

** Gilfillan Bros. Inc., Los Angeles, Calif. 




[J. S. M. P. E. 





(9) Reflection and refraction (caused by glass envelopes). 
(10) Optical systems. 

Image resolution will be discussed in its broader sense, necessarily 
including frame, flicker, and line frequencies which, because of the 
scanning methods used, are some of the direct determinants of reso- 
lution within a given transmission band width. 

The industry and the Federal Communications Commission are 
unanimous in accepting a 6-megacycle channel as the maximum de- 
sirable for the transmission of 
a single public television service 
within the presently useful por- 
tion of the radio frequency spec- 
trum.* Let us, therefore, use 
this as a basis for consideration 
of other factors. 

Fig. 1 shows the manner in 
which such a channel is at pres- 
ent utilized. Part A is the emis- 
sion characteristic of the trans- 
mitters, and part B is the overall 
response characteristic for sight 
and sound sides, in the receiver. 
Sound transmission and recep- 
tion, accomplished at the high- 
frequency end of the channel, 
has very little to offer in the way 
of alternatives except for the 
possible use of frequency instead 
of amplitude modulation. High 
fidelity is practicable with either 
method. The sound side is mentioned here only because of its re- 
lation to the rest of the channel and its consequent effect on remain- 
ing band width available for video information. Modifications in 
the location of sound and video carriers would result in slightly 
increased picture definition, but probably would add to receiver cost 
in an unreasonable ratio. 

In constructing or reconstructing a television image, the function- 
ing of scanning circuits requires a definite portion of time. This 

* From a statement by E. K. Jett, Chief Engineer for the Federal Communi- 
cations Commission at the NTSC organization meeting, July 31, 1940. 


FIG. 1. Typical television channel. 


interests us in considering resolution because, in one sense, it reduces 
the area on which we may resolve a given scene. Fig. 2 gives an 
idea of the relation between the total and useful areas presently em- 
ployed. This illustration is in error principally by the actual beam 
retrace time which, in this instance, is quite small, thus giving the 
desired approximation. The unusable vertical portion is approxi- 
mately seven per cent of the field interval and is approximately 
fifteen per cent of the line interval. These percentages have been 
found highly acceptable as a tolerance within which individual trans- 

FIG. 2. Showing relations between total and useful areas 
now employed. 

mitters and receivers are able to consistently operate. It will be 
noted that percentages of line and field intervals are used here to ex- 
press time, as the actual time will vary with different line and field 

The type of wave-form which initiates the scanning sequence in- 
terests us in considering resolution because we depend upon it for 
accurate interlace and line register. The mind can conceive no end 
of possible wave-shapes and combinations, but only three seem to 
have received much industry attention. The first is that proposed 
by the Radio Manufacturers Association and shown in Fig. 3. The 
chief merits of this system of pulses are that there is considerable 
latitude in receiver circuit design; this type of pulse will control 



[J. S. M. P. E. 




either driven or synchronized scanning oscillators; and has been 
proved satisfactory in use. 

Fig. 4 is an oscillographic reproduction of the RMA field pedestal 
wave-form which may easily be obtained in practice. 

FIG. 4. Oscillographic reproduction of RMA field 
pedestal wave-form. 

Fig. 5 is similarly reproduced, but shows a line pulse and pedestal 
with considerably greater clarity. 

The second method is related to the above in that the same end- 
point is realized, but in a different manner. It has been proposed by 
V. A. Loughran of the Hazeltine Service Corporation that the ampli- 

FIG. 5. Similar to Fig. 4, but showing line pulse and 
pedestal more clearly. 

tude-modulated video signals be interspersed with different fre- 
quency-modulation excursions of the carrier for the line and field 
synchronizing signals. Fig. 6 shows how they propose to accomplish 
this. The advantages are an increase of picture signal equivalent 
to an increase of 60 to 70 per cent in transmitter power, a better 
sync amplitude margin over picture components, a better sync 
signal-to-noise ratio, and no "infra-black" level. 

Fig. 7 is their block diagram of a receiver for this second method. 
The band-pass filters and amplitude selectors would seem to place 



[j. a M. p. E. 

an addedjcost burden tin each receiver purchaser over the first method 
described, to which there might well be some objections. The ad- 
vantages mentioned, or the relative need for them, have not yet been 
confirmed by suitable field tests. 



FIG. 6. Loughran's wave-form for initiating scanning 

The third method is illustrated in Fig. 8, proposed by the Allen 
B. DuMont Laboratories. This latter method seemingly differs from 
the first method in only one essential respect the substitution of a 

FIG. 7. Block diagram of Loughran's receiver. 

train of relatively high-frequency pulses for the broad pulses and the 
elimination of equalizing pulses. 

Another phase of the problem which has wide industry agreement is 
that odd-line interlaced scanning shall be used. This has the ad- 

Jan., 1941] 



vantage of doubling the flicker frequency for a given frame fre- 
quency among other things and will probably be retained even 
in the case of color images. 

Briefly in review, the number of elements composing a single frame 
divided by two, times the frame rate per second, determines the 
band width we have to transmit and subsequently utilize to reproduce 
the image. 

FIG. 8. Signal proposed by Allen DuMont Laboratories. 

In order to get the greatest resolution in a single frame, a decision 
must be made as to the least practical number of frames per second. 
Two major problems are flicker and motional image smear. 

Flicker in television differs somewhat from that in motion pictures 
in that there is continual scanning of lines from top to bottom, in- 
terlaced; each line in the sequence being brighter to the right than 
to the left, and each line in turn diminishing in brightness as another 
line in the scan is added. Special investigations of this type of flicker 
have led to the conclusion that approximately fifty interlaced fields 
(or twenty-five frames) are required to remove the sensation of 
flicker which, as an end point, agrees well with motion picture 
practice. 1 



[J. S. M. P. E. 

Motional smear or blur is not strictly confined to element displace- 
ment due to the progression of an object during the interlaced field 
scanning interval. To evaluate the importance of element displace- 

a & c 

FIG. 9. Motional smear of element detail. 

ment, Fig. 9 shows a group of idealized interlaced lines in which a 
group of tall objects of element width are scanned. In B, the ob- 
jects have moved one-quarter of their width during the field interval. 
It is generally conceded that displacement beyond this amount be- 


441 LIME 30-60 IMTERLflCE X 62SUHE 15-30 IHTERlflCE 


FIG. 10. Object for illustrating horizontal motional smear. 

gins to destroy the fine structure of the image. C shows an extreme 
case where motional element displacement is equal to element width. 
As the total number of elements composing a frame becomes larger 
(i. e., the number of lines increased), each element becomes smaller 
and the rate of permissible motion in a given field interval becomes 
less. Areas of element size may in this case be likened to grain in 


motion picture film. It is proper, therefore, in considering motional 
smear to attach the greatest importance to the image repetition rate 
since element detail serves principally, as illustrated, to define the 
object itself rather than the position of the object in successive field 

It has been established as a physiological fact that the eye sees 
every moving object sharply. The optical axis of the eye stays 
focused on the moving object, and sees no unsharpness (except for 
other motions of parts of the object). 2 Within practical limits, 
sharpness in moving objects is a system requirement. Figs. 10 to 13 





441 LIME 30-60 IMTERLflCE /' 6,25 UME 15-30 IMTERLflCE 








X I 

x "TO 


FIG. 11. Effect at the end of Veo second. 

illustrate the effect, neglecting in the main a slight shape distortion 
of the object. Let us assume the object is the disk shown, and that 
it moves from left to right,, and that the line of transit does not ap- 
preciably change the size of the object, which is approximately one- 
quarter the height of the frame. 

At the present moment, there is some interest in the practicality 
of fifteen as against thirty frames per second, and the questions raised 
are not only on flicker, but on continuity of motion and motional 
smear. 3 Let us, therefore, compare the two. 

With the disk in motion, these systems will scan as shown in Fig. 
11 at the end of Veo second. The 30-frame system has scanned one 
field, and the 15-frame system, half a field. 

74 C. F. WOLCOTT Q. S. M. P .E. 

At Yao second, Fig. 12 shows the 441 -line 30-60 interlaced system 
has completed the scan of one frame. From this point on, its ap- 
pearance will remain unchanged, due to scanning repetition. 

At Vis second, Fig. 13 shows the scanning of one frame also com- 
pleted for the 625-line 15-30 interlaced system. The object has 
travelled twice as far and the difference in smear is evident. What 
happens after this point is that in each system the object "creeps" 
by successive field displacements. 


441 LIME 3O-6O imtKLHCt / 625 LIME 15-30 IMTERLflCt 


FIG. 12. Effect at the end of Vao second. 

A somewhat similar effect would be obtained in the vertical direc- 

In the example given, the rate of motion has been assumed slow 
enough to produce "simple smear." In the 30-frame case the rate 
of motion would have to be twice as fast as in the case of 15 frames 
to produce the greater degree of smear, or that still faster motion is 
required to produce the illusion of multiple or successive images. 

Black-and-white television picture tubes at present used with the 
30-60 interlaced system have what is considered adequate brilliancy 
in a semi-dark room, a relatively fast light decay characteristic, and 
depend on the scanning frequency to eliminate both flicker and 
motional smear. 

Picture tubes have been proposed for the 15-30 interlaced system 
which would have a relatively slow light decay characteristic, the 


intent being to diminish flicker. Assuming equal brilliance and 
color, it would appear that as the tendency toward flicker is di- 
minished, the tendency toward motional smear would be increased, 
and vice versa, within the limits of the system. 

It has been estimated that about one-third of the present 24-frame 
per second motion picture action shots are unnaturally reproduced. 2 

Considering all these factors, thirty frames per second is herewith 
recommended for black-and-white television images. 


tIHE 30-60 IHrKL*Ct ' *25 IMC. 


FIG. 13. Effect at the end of Vu second. 

Before considering resolution in terms of picture detail, let us re- 
view a present system limitation of first importance. The electron 
beam used in scanning (both at the transmitter and receiver) has 
essentially the characteristic shown in Fig. 14. Due to the (approxi- 
mately) cosine-squared intensity distribution in the spot, it is neces- 
sary to overlap the successive scanned lines in order to produce a 
flat field, by a factor which reduces the resolution to an amount 
variously estimated at from 53 to 85 per cent. 4 ' 6 The correct figure 
is apparently in the neighborhood of 70 per cent. 6 

Dr. J. R. Pierce presented a paper "Rectilinear Electron Flow in 
Beams," before the Institute of Radio Engineers Pacific Coast Con- 
vention this fall, which together with subsequent discussion in- 
dicates that square or preferably oblong spot shapes of even intensity 
distribution are at present practical for commercial television camera 

76 C. F. WOLCOTT [J. S. M. P. E. 

or picture tubes. The Bell Telephone Laboratories have constructed 
such a tube experimentally, and the results obtained verify that this 
is a correct approach to the ideal condition. The potentialities of 
this spot shape should be considered in determining standards. 

To emphasize this point further, two illustrations are reproduced 
from a paper by Dr. P. C. Goldmark. 7 Fig. 15 is a direct reproduc- 
tion of the original photograph. When this photograph is optically 
scanned with a slit aperture to simulate a 441 -line 30-frame 60-field 
image transmitted within a 4.25-megacycle band, the result is Fig. 
16. Much of the detail is element size or smaller, critically revealing 
the limitations of this finite scanning spot size, but eloquently testi- 





FIG. 14. Characteristic of electron beam used in scanning. 

fying that television images, as received today, can be improved 
considerably before the maximum resolution of a 441 -line system is 

Table I presents an opportunity for some interesting comparisons. 
In computing the various numerical values shown, some minor as- 
sumptions were made which were the same in all cases, and hence 
serve for comparative purposes. 

We should note parenthetically under the second column, which 
is for 441 lines, the values for vertical and horizontal resolution. 
By cutting down spot size slightly though not enough to depart 
from an apparently flat field as far as the eye can discern resolution 
can be increased to 300 vertically and 400 horizontally. Pictures 
with this degree of resolution have been transmitted on regular pro- 
grams. Corresponding values in the other columns could likewise be 

Jan., 1941] 



increased. This degree of excellence has by no means been an aver- 
age condition during, say, the past year. 

Higher order resolution is basically desirable to all. Those who 
have advocated 625 lines, 15 frames, with the present "cosine- 
squared" spot will doubtless welcome the previously mentioned de- 
velopments, indicating an approach to equivalent results with a 30- 
frame image. 


30 Frame 



















































Number of vertical lines 
3:4 Aspect ratio, H. elements 
Approximate band width in MC 
Useful vert, lines ( 7% blank) 
Useful h. elements (15% blank) 

With Cosine-Squared Spot 

Vertical resolution 
Horizontal resolution 
Approximate band width in MC 

It will be noted that 441 lines is indicated as the maximum which 
might be used with a better spot shape because of the band-width 
limitation. The actual scanning spot shape is a factor in the band 
width required, which may be greater or less than shown; if greater, 
some use may be made of the half megacycle which has been 
allowed for attenuation of video information between video and 
sound channels. This analysis is made on the assumption that ele- 
ment detail is to be continuously transmitted. While this assump- 
tion defines the ultimate excellence of a system, in practice this is by 
no means an average condition. Because of this, the selection of 
507 or 525 lines might be warranted; this to approximate an image 
of the excellence indicated in the top bracket, not present practice. 

Fig. 17 is a photograph of an image derived from a monoscope tube. 
In producing this particular image, a "cosine-squared" spot with 
441-line 30-frame interlaced field scanning was used. This illustrates 
very closely the maximum resolution we can expect under these con- 
ditions and corroborates the values given in the previous figure. 

It will be noted that the line scanning is properly overlapped, 
making it impossible to discern any departure from a flat field. 

It might well be commented on here that some programs received 
in the past with this system (both in the East and West) have, as 
previously noted, been lacking in many respects. At various times, 



[J. S. M. P. E. 

observed definition has been low, conceivably due to receivers and 
their installation or operation; shading has been inadequate; con- 
trast poor, possibly the result of experimental lighting, inadequate 
carrier modulation, or receiver adjustment; variations have been 
noted in synchronizing information; and (to cite a particular case) 
W6XAO has, in the past, been faced with the almost impossible task 

FIG. 15. 

Original photograph from which Fig. 16 was reproduced (Photo by 
Fair child Aerial Surveyor). 

of eliminating vertical bars from the image occasioned by the physi- 
cal proximity of KHJ on 900 kilocycles. W6XAO is now effecting 
a long contemplated change of location to Mount Lee, which will 
eliminate this difficulty. Again speaking in a national sense, some 
people have formed adverse opinions because of program material. 
None of these exprimental difficulties are system limitations. 

Another factor which we must take into consideration is the pos- 
sibility that the public will not, in every case, actually purchase a 
television receiver capable of resolving maximum transmitted detail. 

Jan., 1941 



There are limiting items such as small picture tubes, which, of them- 
selves, would not be capable of maximum definition. In addition 
to this, and perhaps more important, are savings to be effected by 
narrow band pass transformers or filters where the gain per stage is 
higher and less stages are required. The first result of such economies 
is a decrease in resolvable horizontal elements. The pattern shown, 

FIG. 16. Result of optically scanning original of Fig. 15 with a slit aperture 
to simulate a 441-line, 30-frame, 60-field image within a 4.25-mc band 
(Goldmark and Dyer" 1 ). 

and subsequent improvements thereon, can suffer considerable loss 
of horizontal definition (which the eye favors) and still give an ac- 
ceptable ratio of horizontal to vertical resolution. It is submitted 
that this decrease in image resolution should be the result of purchas- 
ing an inferior receiver much as a customer now makes his choice 
between, let us say, an excellent console radio and one of the popular 
midgets rather than a system limitation. 

If, for reasons previously mentioned, it might seem desirable to 
adopt a 507 or 525-line system, the ratio of horizontal to vertical 

80 C. F. WOLCOTT [J. S. M. P. E. 

resolution for less costly television receivers would, for a given band 
width, be much less favorable. This fact must be seriously weighed 
against the advantages sought. 

There are several problems associated with colored images which 
are well illustrated by previous examples given for black-and-white 

In a recently proposed color system, sequential field scanning 
through rotating color filters has been demonstrated.* In com- 
parison with black-and-white images, there are two difficulties which 
can be adequately minimized by a substantially higher frame fre- 
quency for the colored images. To illustrate this need, in the figures 
previously used to show motional smear, we can mentally substitute 

FIG. 17. Photograph of an image derived from 
a monoscope tube. 

alternating and progressing fields of color. Mainly on the trailing 
edge of a moving object, a rainbow effect will result, its degree of 
prominence depending on the rate of motion and the colors of the 
moving object and its background. Also, in comparison to black- 
and-white images, the color flicker frequency is reduced, and for a 
given predominating color, such as a sky background, a higher frame 
rate is indicated. 

Technical details to be overcome include loss of light in the color 
disk, both at the transmitter and receiver. In the receiver, a very 
rapid decay period must be used on the picture tube screen. 

* At the studios of the Columbia Broadcasting System, New York, N. Y., 
September 30, 1940. 


Due to present disparities between black-and-white and color 
systems, two separate standards are indicated. It is suggested that 
the standard for color be set when a system has received adequate 
field tests. Some component parts, both at the transmitter and re- 
ceiver, are not necessarily identical or interchangeable. 

Considering only black-and-white images in summary, and con- 
ceding that the art has by no means developed the full possibilities 
of any system, some latitude for development must, at this time, be 
provided. All presently proposed systems have a common minimum 
flicker and motional smear requirement, indicating for this country a 
30-frame, 60-field, interlaced scanning sequence. The band-width 
limitation is not exceeded under present practice with a 441 -line 
image, but it has been shown that it may well be exceeded as the art 
progresses. A possible compromise is indicated in the use of a 507 
or 525-line image. 


1 ZWORYKIN, V. K., AND MORTON, G. A. i "Television the Electronics of 
Image Transmission," J. Wiley & Sons (New York), 1940. 

2 LEHMAN, F.: "Considerations on Optical Rectification," Die Kinotechnik, 
14 (Dec. 5, 1932). 

3 "You, Too, Can Have Television," a pamphlet, Allen B. DuMont Labora- 
tories, Inc., Passaic, N. J. 

4 MERTZ, P., AND GRAY, F. : "A Theory of Scanning and Its Relation to the 
Characteristics of the Transmitted Signal in Telephotography and Television," 
Bell Syst. Tech. J., 13 (July, 1934), p. 464. 

6 KELL, R. D., BEDFORD, A. V., AND FREDENDALL, G. L. : "A Determination of 
Optimum Number of Lines in a Television System," RCA Review, 5 (July, 1940), 
p. 8. 

6 WHEELER, H. A., AND LOUGHRAN, V. A.: "The Fine Structure of Television 
Images," Proc. I. R. E., 26 (May, 1938), p. 540. 

7 GOLDMARK, P. C., AND DYER, J. N. : "Quality in Television Pictures," /. Soc. 
Mot. Pict. Eng., XXXV (Sept., 1940), p. 234. 


OF 1940* 


Summary. This paper describes the methods by which the National Broadcasting 
Company provided television coverage of the Democratic and Republican political con- 
ventions, for the benefit of the television audience in New York City and surrounding 

Due to lack of television transmission facilities between Chicago and New York, 
portions of the Democratic Convention were recorded on film. The film was sent to 
the NBC film scanning studio in Radio City and presented as part of the afternoon and 
evening television programs. 

Portable television pick-up equipment was sent to Philadelphia for the purpose of 
televising the Republican Convention proceedings. The television signal was trans- 
mitted 104 l /z miles between the pick-up point and the NBC television transmitter in 
New York City by means of the coaxial cable facilities of the Bell System. 

The National Broadcasting Company inaugurated regularly 
scheduled experimental television programs in the New York City 
area on April 30, 1939, when its television mobile unit cameras oc- 
cupied a space on the camera platform at the Court of Peace, New 
York World's Fair. Thousands of persons in their homes miles away 
from the Fair Grounds, at Flushing, Long Island, were able to see and 
hear President Roosevelt dedicate the New York World's Fair of 1939. 
Between that date and January, 1940, these television mobile unit 
cameras were present at over ninety events of public interest occur- 
ring within a radius of twenty-five miles of the Empire State Building, 
where the receiver for relay work was located. Practically every 
phase of sports as well as most of the major news events occurring in 
New York City and its suburbs has been televised during these eight 

The program management of NBC looked forward to the coming 
political conventions as a source of interesting television program 
material. It was hoped that both major parties would choose New 

* Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received Novem- 
ber 7, 1940. 

** National Broadcasting Co., New York, N. Y. 


York City as the site of their national conventions, and thereby make 
available to the television audience two of the most important do- 
mestic news stories of the year. They were mindful of the fact that 
only sixteen years ago, the then budding young radio industry had 
achieved great success at the Democratic Convention of 1924 when 
the vote of the Alabama delegation of "24 votes for Underwood," 
had echoed for days in New York's old Madison Square Garden, and 
had been heard by a few thousand early radio set owners as they 
passed around the head-phones. 

The choosing of Philadelphia by the Republican National Com- 
mittee early this year as the site of their Convention and that of 
Chicago by the Democratic National Committee for theirs, made the 
possibility of a direct television pick-up by NBC of a national politi- 
cal convention possible in only one case. The relay transmitters in 
use for field transmission could not have spanned the 90-mile distance 
between New York and Philadelphia, and other relay links under con- 
struction could not have been made available in time. The number 
of television relay transmitters required to establish a 900-mile cir- 
cuit between Chicago and New York could not have been assembled 
and tested in less than two years. A coaxial cable installed a few 
years ago between the Bell Laboratories Building at 32 Sixth Avenue, 
New York, and the Bourse Building at 4th Street, Philadelphia, was 
the only means of a television relay. This cable had been installed 
for the experimental study of multicarrier telephony. The original 
band-pass characteristic of its associated repeaters had been estab- 
lished at 1000 kilocycles. Such a frequency band is not sufficient to 
convey a television image of good definition. Inquiries made of the 
Bell System engineers revealed that they were at work on a planned 
extension of the band-pass characteristic to 2800 kilocycles. An 
amplifier having such a frequency characteristic and minimum phase 
delay is capable of transmitting sufficient of the video frequencies 
comprising a 441 -line television picture to result in an image of good 

A survey was made at Convention Hall to determine the practi- 
cability of adapting the NBC television field equipment to the use of 
the hall planned by the arrangements committee. Tentative permis- 
sion was secured to televise the proceedings of the Convention. The 
engineers of the Bell System welcomed the opportunity of putting 
this cable to a practical test as the first television cable network link 
in the United States. 

84 H. P. SEE [j. s. M. P. E. 

The first problem confronting the engineers engaged in this pio- 
neering attempt was establishing some method of connection between 
the New York terminus of the cable at 32 Sixth Avenue, and the Radio 
City television control room, which are six miles apart; and the con- 
nection between the Bourse Building at 4th Street, Philadelphia, 
and the Convention Hall at 32nd Street, Philadelphia. The latter 
points are three and one-half miles apart. It was originally planned 
that the two ultra-high-frequency transmitters operating in conjunc- 
tion with the regular video equipment used in the New York City 
area would be used to span these gaps. This plan was rejected in 
favor of an attempt to use the experience gained in experimenting 
with specially selected highly equalized pairs among the regular 
underground telephone facilities of the Bell System. Several pro- 
grams had been transmitted between Madison Square Garden and 
the television control room on the fifth floor of the RCA Building in 
New York. This distance is slightly less than two circuit miles. 
Through the use of the equalizers and repeaters a good degree of suc- 
cess had been achieved. 

Two of these twisted-pair circuits were set up with repeaters at 
one-mile intervals between Convention Hall and the Bourse Building 
and between the Bell Laboratories Building and Radio City. The 
entire circuit was electrically adjusted and tested by the Bell System 
engineers over a period of two months. Portable television pick-up 
equipment was shipped from Radio City to Philadelphia and used for 
test purposes. The first test picture was transmitted between these 
two eastern seaboard cities on June 14, 1940. With a cable connec- 
tion established between the Convention Hall and Radio City it re- 
mained for NBC to use its limited amount of field equipment to the 
best advantage in covering this news event side by side with radio 
broadcasting and newsreel facilities. 

During the preliminary surveys made at the Philadelphia Conven- 
tion Hall, committee members in a sincere desire to coordinate facili- 
ties, suggested that television cameras occupy a space on the news- 
reel camera platform. This suggestion was taken under advisement 
but in view of the dissimilarity of equipment adaptation and the di- 
vergent modes of operation of the newsreel and television systems, it 
was decided that a special television platform should be constructed. 
It has often been freely stated that television is a marriage of radio 
broadcasting and motion pictures. While this may seem to be true 
in a broad sense, and in the mind of the television viewer as he watches 


scenes occurring at points remote from him, the practical operating 
requirements at the present state of the television art, and especially 
in field work, are of necessity quite different from those of the news- 
reels. These differences were very well exemplified at the Philadel- 
phia Convention. It is, of course, quite probable that the furtherance 
of the television art and improvements in pick-up tubes may change 
the complexion of this matter especially as regards optical systems, 
but the modes of operation will remain basically different. 

The television camera at the present time must be located in such a 
manner that it achieves the greatest possible radius of action from 
one position. It is, for all practical purposes, tied down to one posi- 
tion by its cable connection to the control equipment. This is quali- 
fied to some extent when the location is in an open space and sufficient 
slack cable may be left to allow movement of the camera on a movable 
platform, or a change in its position while another camera takes over 
the operation. Under conditions such as existed at the Convention 
pick-up, where the camera had to be located on a platform, little or no 
change in position is possible. 

There are three television pick-up tubes commercially available. 
These are the large Iconoscope, the small Iconoscope, and the Orthi- 
con tubes. The large Iconoscope sensitized surface is 4.81 by 3.62 
inches, the small Iconoscope 3 by 2.75 inches, and the Orthicon 2.33 
by 1.74 inches. All three tubes are in use in NBC television field 
equipment. A large Iconoscope camera and an Orthicon camera are 
used with the Telemobile Unit, while two small Iconoscope type cam- 
eras are used with the transportable equipment. 

While the operating sensitivity of a pick-up tube may be affected by 
a number of factors and a direct quantitative comparison of tubes is 
difficult to make, it has been found that the Orthicon tube is more sen- 
sitive in practical operation than either of the other two types. This 
is particularly evident when the overall brightness of the scene is low. 
The characteristics of this tube are explained in the October, 1939, 
issue of RCA Review and in the July, 1939, issue of Electronics. 

In consideration of the fact that operation at Convention Hall was 
to take place under artificial illumination, the Orthicon represented 
the only tube usable under these conditions by which close-up pic- 
tures from a distance could be obtained with commercially available 
optical systems. 

From the dimensions previously given, it is seen that the sensi- 
tized surface of an Orthicon tube is much greater than that of the 35- 

86 H. P. SEE tf. S. M. P. E. 

mm film. In order to obtain an image size proportionate to the 
area, a television camera equipped with an Orthicon tube must use a 
lens which has approximately 2.7 times the focal length of that in use 
in a newsreel camera when the two cameras are located at the same 
distance from the subject matter. If the lens speed is to be retained 
under these conditions, the lens diameter is also increased by a factor 
of 2.7. In most lenses suitable for our purposes, these considerations 
involve lens diameters of magnitudes which are not commercially 
available and whose prohibitive cost for special construction render 
them economically unfeasible. Lenses may be constructed in certain 
large-diameter ranges without correction for chromatic aberration. 
The use of such a lens would, of course, leave much to be desired. If 
the lens diameter is not increased, a loss factor of approximately 7 
in mosaic illumination will result when the focal length is increased 
by a factor of 2.7. It is therefore evident that a television camera 
using the smallest mosaic area now available should seek a location 
closer to the subject matter than is necessary for a newsreel camera. 

At the present time, a camera capable of housing the Orthicon tube, 
magnetic focusing coil and deflection coil, video and deflection am- 
plifiers, is 28 inches long. When this camera is located at a distance 
of 50 to 60 feet from a speakers' rostrum, a 20-inch focal-length lens is 
the minimum that may be used in order to achieve a semi-close-up 
view of the average man. The addition of this lens to the camera re- 
sults in a unit 4 feet long. It is inconceivable that such a unit could 
be crowded together with five newsreel cameras on an average-size 
camera platform and still not be impeded or interfere with the other 
cameras when panned in an arc of 60 degrees. 

Film editing is done in a matter of hours after the scene recorded has 
passed into oblivion and become history. By accepted methods, 
scenes may be taken in any sequence, sound added, and the final re- 
sult shows the highlights of an event in a few short moments. It is 
therefore possible to show many varied scenes and yet not have used 
more than one or possibly two cameras. 

Television editing is a matter of instantaneous decision, and is 
achieved by electrically switching to another camera system, which 
may be focused on the previously shown scene but using either a 
wider or narrower lens system. It may also be a camera located at 
some distance away from the first one used. Television operation is 
a continuous proposition, and therefore more facilities must be avail- 
able to maintain varied coverage of a particular event. Mistakes in 


the motion picture industry are found in the cutting room while mis- 
takes in television operation are seen in the living-room. 

In order to have covered the Convention in a manner comparable 
to the combined facilities of the radio broadcasting and newsreel 
companies, we should have found it necessary to place a minimum of 
six cameras at the Convention Hall and two or three more at the 
various headquarters of political groups in the city. 

The Convention was the first indoor event wherein newsreel and 
television cameras were to be focused on the same action over a period 
of hours and days and yet maintain diametrically opposed visual 
coverage from one minute to the next. Lighting at this Convention 
therefore had to be considered from the television standpoint as well 
as that of the newsreels. 

As distinguished from newsreel operation, television operation is a 
continuous performance and thus the lighting for television must be 
continuously maintained. It is possible for newsreel camermen to 
make the best use of their flexibility in editing by filming the keynote 
address or at least the highlights of this speech on the day before the 
opening of the Convention. This arrangement is made with the key- 
noter. The cameras are set up on the floor below the speakers' ros- 
trum. When the keynote speech is actually given, during the Con- 
vention, the newsreels may have the light units in the hall spread 
upon the seated delegates, and make "shots" of the audience reaction 
to the keynote address. This is the first glimpse that television has 
had of the keynote speech, and the speaker must be adequately il- 
luminated for the television cameras. In like manner, during nomi- 
nating and seconding speeches, the newsreels may, if they desire, 
take only excerpts of even the most important speeches. Armed 
with advance script, they may take only those portions in which they 
are interested. These are the highlights. For such operation, it is 
not necessary that continuous high-level lighting be maintained. 
For television purposes, a threshold of illumination must be main- 
tained on the speakers' rostrum at all times. 

Sufficient lighting for television was maintained at Philadelphia by 
increasing the specifications for fixed lighting used in the same hall in 
1936. Ten 5-kw solar spotlights were suspended from the ceiling and 
two similar 5-kw units were located on each balcony. The overhead 
lamps were mounted on a frame above and in front of the speaker. 
This source of light provided an illumination of 800 foot-candles in- 
cident on the speakers' platform. 

88 H. P. SEE [J. S. M. p. E. 

Eight 150-ampere d-c arc lamps were installed at points on the 
balcony so that they could be directed to any spot on the floor for 
both newsreel and television purposes. These lamps were used to 
illuminate an area on either side and in back of the speakers' rostrum 
when the newsreels wished extra illumination in that vicinity. When 
the center of interest was transferred from the speaker to the state 
delegations, such as during balloting for candidates, the television 
cameras followed the beam of the arc lamps. If the television pro- 
gram director desired that the cameras be focused on the speakers' 
rostrum while the newsreels were still shooting scenes of the floor, the 
threshold of illumination provided by the fixed lights focused on the 
speakers' rostrum was still maintained for television operation. 

FIG. 1. Unit 1A parked in front of Radio City on occasion of 
sidewalk interviews. 

The television field equipment available for use at the Philadelphia 
Convention consisted of Telemobile Units 1A and IB and the trans- 
portable, or suitcase type of equipment. Telemobile Units 1A and 
IB are each 10-ton motor vehicles and are 26 feet long. Telemobile 
Unit 1A contains the video and audio equipment which operates in 
conjunction with one television camera employing a standard Icono- 
scope pick-up tube and one camera using an Orthicon tube. The 
equipment, which consists of a synchronizing generator, deflection 
amplifiers, control amplifiers, line amplifiers and monitors, and audio 
system is mounted on nine racks situated on the centerline of the 
vehicle in the long dimension. All cables and accessories necessary 
for the operation of this unit are carried within the body. Telemo- 
bile Unit IB contains a 400-watt ultra-high-frequency relay trans- 
mitter licensed to operate within the channel 162-168 me monitoring 
equipment, and cable accessories (Fig 1). 



The transportable, or suitcase type of equipment, consists of eleven 
boxes of an average weight of 50 pounds each. When interconnected, 
this equipment provides two television cameras employing the small 
type Iconoscope, a synchronizing generator, amplifiers, monitors, and 

FIG. 2. ( Upper} Orthicon camera on camera platform 

of Convention Hall. 

FIG. 3. (Lower} Another view of cameras at Con- 
vention Hall. The Iconoscope camera is at the left. 

a low-powered ultra-high-frequency television relay transmitter li- 
censed to transmit in the channels 282-288 me and 288-294 me. 

The Telemobile Units were constructed in 1937 and placed in op- 
eration in 1938. The transportable equipment was placed in opera- 
tion in the fall of 1939. Several distinctive features which are the 
result of improved design at present prevent the cameras of one set 
of equipment from being directly connected to the video and control 
apparatus of the other. 


H. P. SEE 

[J. S. M. P. E. 

The provision of an all-cable circuit between the pick-up point and 
Radio City, as previously outlined, obviated the necessity for con- 
sidering the radio-frequency transmitters as part of the equipment 
necessary for transmission from Convention Hall. The low-power 
transmitter was held in reserve together with its associated receiver, 

FIG. 4. Interior of telemobile unit 1A parked on 
freight-loading ramp at Convention Hall. View 
shows engineers and producer. 

to provide for any contingency, such as the possibility of an accep- 
tance speech being delivered by the party's candidate at a large out- 
door stadium nearby. In that event, this transmitter would have 
provided the link to the cable termination at Convention Hall. 

The purpose of televising the proceedings of the Convention was to 
bring to the television audience the greatest possible amount of pro- 
gram material on as wide a scope as was consistent with equipment 
and personnel limitations. The Main Hall, in which the 1000 or 


more delegates and their alternates were seated under their identify- 
ing State banners was the focal point of interest as speaker after 
speaker mounted the rostrum to address the assembly. The main 
entrance to the Hall was another point of interest as the crowds 
gathered to watch nationally known figures disembark from taxis 
and enter the auditorium for each of the many sessions occurring dur- 
ing the five-day Convention. The possibilities of providing a space 
where news commentators, political figures, and others might be in- 
terviewed before the television camera, was considered as an impor- 
tant addition to the existing program sources. 

A television camera platform was constructed on the south balcony 
of the main arena at a distance of 60 feet from the speakers' rostrum. 
The platform was 8 by 5 feet and was designed to accommodate two 
cameras, two camera operators, and an announcer. The camera 
faced the speakers' rostrum at an angle of approximately 45 degrees, 
and looked down at it from an angle of approximately 10 degrees. 
The position of the platform was predicated upon the already chosen 
location of the newsreel camera platform and a compromise between 
our distance from the speaker and the angle at which the camera 
faced him. The newsreel platform was placed at the same propor- 
tionate position that it had occupied during the Democratic Conven- 
tion held at the same hall four years ago. The television platform 
was 15 feet forward from the newsreel platform. With both these 
platforms built out from the balcony on the speaker's left, the 15- 
foot dimension allowed sufficient spacing to prevent the television 
camera platform and equipment from obstructing the view of the 
newsreel cameramen located nearest the balcony when he was using a 
30-degree angle lens with his equipment (Figs. 2 and 3). 

The two cameras associated with the Telemobile Unit were mounted 
side by side on this platform. A 19 3 /4-inch focal length //4.5 mag- 
nesium fluoride treated lens was used on the Orthicon camera. The 
angle of view included a space approximately 3 feet either side 
of the speaker. When viewed on a 12-inch Kinescope, a bust view 
of the speaker occupied 5 inches vertically. The Orthicon was also 
used to pick up floor scenes during the balloting and demonstration 
parades which followed the placing of a candidate in nomination. 
The Iconoscope camera was equipped with an 8-inch focal length 
f/2 treated lens. The wide angle of this lens included almost the en- 
tire width of the stage when focused on the speakers' stand. When 
focused on the floor, this lens showed more than half of the State 


H. P. SEE 

[J. S. M. P. E. 

delegations. The sensitivity of the Orthicon enabled it to be used 
satisfactorily on scenes having a lower light level than used on the 
speakers' stand. 

The large mobile unit containing the video equipment necessary 
for the operation of the main hall Iconoscope and Orthicon cameras 
was parked on a freight-loading ramp one floor below the main arena 
and inside the Hall proper. Circuit constants of the equipment and 
other factors limit the distance possible between the cameras and the 
mobile unit to a length of camera cable not exceeding 250 feet. Spe- 

FIG. 5. Transportable type camera used in specially constructed studio. 
View shows Mrs. Wendell Willkie being interviewed by Mr. A. H. Morton, 
NBC Vice-President in charge of television. 

cial holes were cut in the concrete floor in order to provide a path for 
the camera cable which would not exceed that limitation (Fig. 4) . 

A television studio was constructed on the second floor of Conven- 
tion Hall. An unused dressing room was converted into a studio for 
interview purposes and a control room adjoining it. The transport- 
able equipment was located in this control room. One of the 
small type Iconoscope cameras was used in the studio. The studio 
was located so as to be readily accessible to staircases, elevators, the 
speakers' platform, and the State delegations. Three 5-kw portable 
lighting units were used to provide illumination in the studio. Air- 
conditioning apparatus was installed to overcome the heat generated 
by these lights and the warm June weather outside. The control 


room was left windowless in order to provide an ideal condition for 
viewing pictures on the Kinescope monitor. The other small camera 
was set up at the main entrance to Convention Hall which is 750 feet 
from the studio and its control room. This position was used during 
the daylight hours to show activities occurring there when there was 
little or no business being transacted in the Convention Hall. It 
was always used at the opening of each program (Figs. 5, 6, and 7). 

In addition to televising the Convention and transmitting it to the 
television audience in New York State and surrounding areas of New 
Jersey, Connecticut, and Pennsylvania, provisions were made to 
satisfy the overflow of the general public which could not gain ad- 
mittance to the Main Hall. The RCA Victor Company installed 
sixty television receivers in a large room set aside for public viewing. 
Receivers were also installed at the temporary Convention Hall offices 
of the three leading national press associations and a Press Club Room. 
The picture signal and sound to these receivers were transmitted by 
means of cable. 

Over two miles of wire were installed in Convention Hall by NBC 
to supply power necessary for operating the television mobile units, 
power for lighting, air-conditioning, and teletypewriters in the studio 
and office, video circuits, audio circuits, sound monitoring, and com- 
munication circuits. This wiring was exclusive of those facilities 
which were installed by the Telephone Company within the building 
for our purposes. Such circuits included the audio connections be- 
tween the mobile unit and the main frame, morse circuit, private line 
connection from the mobile unit to the Telephone Company's video 
control equipment in another part of the building and an ordinary 
business phone circuit. A breakdown of our temporary wiring in- 
stallation shows that 2155 feet of coaxial cable and 1250 feet of camera 
cable were used. The remainder is divided between communication, 
audio, and power wiring facilities. These figures are included in or- 
der to indicate the magnitude of a temporary television pick-up at an 
event of this character. Many of our regular routine pick-ups in 
New York City at which permanent wiring facilities do not exist 
necessitate the temporary installation of close to one mile of wiring 
for a one-hour program. 

It has been previously mentioned that differences in the equipment 
do not permit the ready interconnection of either pair of cameras with 
the corresponding control equipment of the other. In order to co- 
ordinate these cameras and provide smooth continuity of the program 


H. P. SEE 

tf. S. M. P. E. 

as the points of interest changed from the program standpoint, it was 
necessary that a rapid switching system be devised whereby the video 

FIG. 6. (Upper} Transportable control room. View 

shows all control equipment. 

FIG. 7. (Lower} Scene showing transportable type 
camera in use at Main Entrance to Convention Hall. 

signals from the transportable and mobile unit equipment could be 
transmitted to the NBC control room in New York City via the co- 
axial cable as well as to the special receiver monitor positions in other 
parts of the building. 


The complete video signal contains picture signals, blanking pulses, 
and synchronizing pulses. Both the transportable and mobile unit 
equipment are complete units, and their outputs contain the signals 
referred to. In addition, it must be borne in mind that each set of 
equipment had its own synchronizing generator. These two systems 
were, of course, not identically synchronous with each other. Had 
it not been for program commitments in New York which kept the 
mobile units busy up until a short time prior to the Convention, there 
would have been nothing to prevent a system's being devised whereby 
the basic pulses from one generator could be used to control the other. 
In standard RCA television equipment, of which the Telemobile Units 
and the transportable equipment are an example, the video signal and 
blanking pulses are fed to an input tube in the line amplifier. The 
synchronizing signal is fed into its own input tube in this amplifier. 
The resultant complete signal is the television signal .which is trans- 
mitted. In the case of this installation the main feed to the Tele- 
phone Company was transmitted from the line amplifier in mobile 
unit 1A . The monitors in the Telemobile Unit received their signal 
from a separate output tube in the line amplifier. This signal was 
also transmitted to another line amplifier which in turn fed two cir- 
cuits. One circuit went to the sixty receivers in the public viewing 
room, and the other to the three receivers in the Press Offices and the 
single receiver in the Press Club Room. A switching system was 
devised so that the total output of the transportable equipment lo- 
cated on the second floor of Convention Hall was available at the in- 
put of the main line amplifier at the mobile unit so that this amplifier 
could receive either the outputs of the mobile unit cameras or the out- 
puts of the transportable equipment cameras as represented by the com- 
bined transportable equipment signal. In this manner a four-camera 
pick-up was obtained. The ratios of synchronizing signals and 
blanking pulses in the mobile unit and in the transportable equipment 
control room were established by engineers in communication with 
each other by private line, so that the overall ratios transmitted to 
the Bell System, the local monitors, and the special monitoring posi- 
tions were identical. 

The audio system for television at this Convention presented little 
or no problem. The three major networks had combined their re- 
sources to obtain pick-ups from the fifty-three floor positions avail- 
able for each of the state delegations. Each network provided its 
own pick-up at the speakers' rostrum. The network feed from all 

96 H. P. SEE [j. s. M. P. E. 

these points was suitably attenuated and transmitted to one mixer 
position on the audio panel in Telemobile Unit 1A. Microphones 
were placed on the camera platform in the main hall, in the television 
studio, and at the street position. 

Telemobile Unit 1A was designated as the master control point 
for the entire system. Three television engineers and a program di- 
rector constituted the personnel at this point. One engineer was lo- 
cated in the control room associated with the studio. Engineers 
were at each camera position. The private line communication fa- 
cilities necessary to interconnect all these points, plus providing com- 
munication between the Bell System control room, the NBC network 
control room, the public viewing room, and the mobile unit surpassed 
any communication installation set-up to "date for either radio broad- 
casting or television. The master system was subdivided so that the 
video engineers and camera engineers at the mobile unit were able to 
converse without interference from conversations going on between 
video engineers and camera engineers at the transportable equipment. 
On a separate ear-phone, however, calls could be established between 
these two circuits after a switch was thrown. The same sort of sys- 
tem was carried through in the case of the audio engineer to the pub- 
lic viewing room, etc. The program director at the mobile unit had a 
separate communication circuit to his assistants at the camera posi- 

The announcing technic in television is considerably different from 
that practiced in radio broadcasting. It is customary in sound broad- 
casting for an announcer to maintain a running comment, and bring 
to his listeners, through words, that which they can not see. In tele- 
vision, the running commentary is unnecessary. In the main hall 
the sequence of events on the speakers' stand and among the dele- 
gates not only predetermined the sound coverage but in many cases 
dictated the movements of the cameras. The only time that the 
announcer in the main hall was heard was during lulls, or when 
events occurring within the range of the camera needed some special 
explanation. The program director, located in the mobile unit, was 
able by means of the monitor system, to see the scenes on which both 
cameras were focused. According to his discretion, and predeter- 
mined program plans, he directed both television camera operators 
and the announcer, when the order of events did not logically prede- 
termine both the visual and aural coverage. 

Perhaps the most rapid-fire television editing by means of elec- 


trically switching cameras equipped with different lens systems oc- 
curred when ballots were taken. As the Chairman called the name 
of the State delegation, he was seen by means of the Orthicon camera. 
A switch was then made to the Iconoscope camera equipped with the 
wide-angle lens. This camera was focused on the general floor scene. 
As soon as the switch was made, the Orthicon camera operator panned 
his camera to pick up a close-up scene of the State delegation. As 
soon as this had been brought into focus it was switched to the line 
amplifier, etc. By means of the intricate communication facilities, 
camera operators, directors, announcers, and video control engineers 
were at the peak of their cooperation during these balloting periods. 

In addition to the communication facilities, it is NBC practice to 
provide a head-set at each point remote from the main control posi- 
tion by which members of both the program and engineering staffs 
are able to hear the sound transmission accompanying the picture. 
This system is known as "feedback monitoring." A loud speaker 
system in the mobile unit is used by the video engineers and program 
directors. By means of this monitoring system, the announcer may 
lead the camera by his description and other expressions which will 
indicate the content of the scene to follow. 

While the position of all our facilities could not be determined until 
the Convention Committees had formulated all their plans, it was 
not possible for the Bell System engineers to delay their additional 
installations and tests necessary before the cable would be usable. 
They started their work in April. The special twisted-pair circuit 
was terminated in the Telephone terminal room on the northwest 
corner of the Convention Hall Building. At a later date, when the 
position of the studio, camera platform, and other factors had deter- 
mined the location of the mobile units, a connection had to be made 
between the terminal room and the mobile unit; 750 feet of flexible 
coaxial cable was needed to complete this installation. 

The completed picture signals from all four cameras were trans- 
mitted to the terminal room over the 750 foot piece of cable. At the 
terminal room the Bell Laboratory engineers were equipped with a 
monitor, an amplifier, and a video equalizer. Two special twisted- 
pair circuits had been installed between the Convention Hall and the 
Bourse Building. There were two repeater stations on these cir- 
cuits. One was at the University of Pennsylvania, and the other at 
the Spruce Central Telephone Office. These picture circuits and re- 
peaters were capable of transmitting a band of frequencies from 30 to 

98 H. P. SEE [j. s. M. P. E. 

2,800,000 cycles per second. At the Bourse Building, a carrier system 
was used. The video signal was transposed so that the lowest fre- 
quency being transmitted through the coaxial cable and the twenty 
repeaters between Philadelphia and New York was of the order of 
300,000 cycles per second. At the New York terminal of the cable, 
this signal was restored to the normal video range, and placed on two 
separate twisted pair circuits and transmitted simultaneously to the 
NBC television control room in Radio City. One circuit was of the 
same type as that used between Madison Square Garden and Radio 
City, and the two between Convention Hall and the Bourse Building. 
This circuit required repeaters at approximately one-mile intervals. 
They were located at the Watkins, Longacre, and Circle central offices. 
The second circuit was a special shielded cable undergoing tests at 
the time. Intermediate amplifiers were not required on this circuit 
even though its length was approximately 4 miles. In all, the pic- 
ture signal travelled 104V2 miles between the pick-up and the Em- 
pire State Transmitter (Fig. 8). 

The combined facilities of NBC and the Bell System were used for 
a total of 33 program hours during the 5-day Convention. Persons 
and events heretofore known to many and yet seen by few were made 
available in close-up upon the many television receivers in the New 
York area. The Convention's candidate became the first presiden- 
tial nominee to be seen by a television audience as he acknowledged 
the greetings of the delegates. 

Faced with the lack of either radio-frequency or coaxial cable relay 
facilities between Chicago and New York, all thoughts of the direct 
pick-up method of bringing the proceedings of the Democratic Con- 
vention to our television audience had to be abandoned. The origi- 
nal consideration of obtaining special motion pictures at each ses- 
sion of this Convention and transporting them by air express to New 
York for television film scanning was adhered to. 

The New York television audience was provided with scenes of the 
first day's activities of the Democratic National Convention held at 
Chicago, June 15th-18th, within 24 hours of the beginning of the 
first session. Special newsreel releases were made up for NBC by a 
newsreel company. At the end of each day's session, these newsreels 
were flown to New York. Developed and edited the next morning, 
they were available for film scanning as a part of the regular after- 
noon NBC television programs. The same releases were shown on 
the regular evening television programs. 



While this method of coverage was not immediate, and lacked some 
of the spontaneity felt by televiewers who looked on the Philadelphia 
Convention, several scenes were available which had no parallel at 




, y 








\7 V 







| | LJ" | 


FIG. 8. Diagram showing circuit routing of television 
signal from Convention Hall to Empire State Building. 

the Philadelphia pick-up. The motion pictures included not only 
scenes within the Hall, interviews with nationally known political 
figures, and street scenes, all of which were provided by the television 

100 H. P. SEE 

cameras at Philadelphia, but included pictures taken at the head- 
quarters in Chicago of the many potential candidates, the arrival of 
colorful figures at railroad stations and other "atmosphere shots." 
The mobility of a small number of motion picture camera units was 
evident. This is, of course, in direct contrast to the lack of flexi- 
bility of the television cameras at Convention Hall, and bears out the 
contention made in the early part of this paper that from six to ten 
television cameras would have been necessary at Philadelphia in 
order to have been able to present the wide and varied coverage of an 
event of this character in like manner to that of the newsreels. 

It is estimated that the television audience witnessing the scenes 
attendant to both the Republican and Democratic National Conven- 
tions of 1940 numbered in excess of 40,000 persons scattered over an 
area of 10,000 square-miles. Persons located over 300 miles distant 
from Philadelphia were able to see and hear events happening in that 
city through the combined facilities of the NBC television field 
pick-up equipment, the Bell System coaxial cable between Philadel- 
phia and New York, the NBC television transmitter atop the Em- 
pire State Building tower in New York City, and the special receiving 
equipment and television transmitter of the General Electric Com- 
pany, located just outside the city of Schenectady, N. Y. The latter 
station rebroadcast most of the NBC programs in the Schenectady 
area. The Philadelphia transmission, which was the 175th NBC 
outside television pick-up since April, 1939, was the forerunner of 
future television network operation and provided the earliest tele- 
vision cable network facilities in the United States with their first 
practical operating test. 



Summary. The paper is an attempt to reduce to words a portion of the mechanical 
and artistic elements involved in the process of editing a motion picture. The authors 
realize that they are dealing with a highly controversial subject but feel that, as there is 
so little pertinent material available on this phase of motion picture production, this 
paper may serve as a preliminary to a study on a larger scale. 

Consideration is given to the origin of film editing and its advancement from the 
purely mechanical craft of the early days to its present status as a contributing factor 
in the entertainment and dramatic values of the motion picture of today. 

It is a far cry from the old nickelodeon theater of the early nineteen 
hundreds to the modern, streamlined motion picture palaces of 1940. 
The business of editing pictures, which is now an important phase of 
the industry, was at that time unknown. In the beginning, it was the 
duty of the cameraman to screen the negative, cut off the number 
plates, arrange the film in continuity, and have it printed at the 
laboratory. This work necessitated long hours for the cameraman, 
and damaged the negative by running it through the early type of 
projector. The "cutter" was first introduced to the film industry as 
a young man whose duty was to assist the cameraman by cutting off 
the scene-number plates and prepare the negative for printing. Be- 
cause the boy used a pair of shears to snip out fogged pieces of nega- 
tive and to remove the scene-number plates, he became known as a 
"cutter." In the early productions, all the scenes were long shots, 
requiring only a few narrative titles identifying the locale as Niagara 
Falls, for example, and telling the audience that Mary and Joe were 
on their honeymoon, etc. 

Competition among the producers was about as keen in those old 
days as it is now, and each was staying awake nights trying to think 
up new ideas for improving his product. One enterprising producer 
conceived the happy idea of having his pictures talk. He used a 

* Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received October 12. 

** RKO Radio Pictures, Inc., Hollywood, Calif. 



talking machine next to the screen, mechanically driven by the pro- 
jector, but this proved impracticable. Another producer, not to be 
outdone, decided to make his actors speak dialog, and so developed 
the use of "spoken" titles. This advancement brought the cutter 
into his own, requiring him not only to write and intelligently edit 
the spoken titles, but thereafter to "cut" them into the proper place 
in the scenes. 

As the film story-telling technic developed, the responsibilities 
of the cutter increased. The industry continued to advance, pro- 
gressing through many stages: the early short subject, purely action 
and pantomime; the two-reel short, with pantomime and narrative 
titles; the first feature pictures of about three reels with spoken 
titles; the five-reeler, with more elaborate sets; then the filming of 
full-length plays and novels; after that came the "super-colossals." 
The cutter had been learning and progressing, step by step, with the 
industry. But now came sound, and with it an entirely different 
technic was demanded of him. The actual piecing together of film is 
generally taken for granted by the public. . .the mechanics of setting 
the stage the introduction of the characters their purpose in the 
scene and their relation to each other. For example, the opening of 
a scene might go like this: 

First, there would be a long shot of a country railway station so 
we know we are not in the city, or China, or on a lake. Then comes 
the close-up of the name of the station Glens Falls, N. Y. Over this 
we can hear the distant sound of a train approaching. Now we know 
we are in the country. From the Glens Falls, N. Y., sign, we pan 
down to a full shot of an old family horse and carriage. A small boy 
and a man are seated in the carriage. A close shot, just holding the 
boy and the man, finds the boy saying: "Well, here comes the train, 
Dad." This is done to tell the audience the sound they hear is a 
train coming and that the man is the boy's father. Now we want to 
see what the father looks like, so we show a close-up of him. He noti- 
fies us that the train is on time and he is sure that the boy's sister 
will be glad to be home from college. Now a fairly long shot showing 
the train coming to a stop. A closer shot shows a pretty girl standing 
on the platform. This would be Lana Turner, waving a college pen- 
nant in one hand and her sweater in the other. 

While on the subject of close-ups, it might be well to discuss this 
most misused piece of film. One can find too many close-ups of the 
entire cast scattered throughout every scene in any show. When one 


inquires the reason, the answer is something like this: "Rhythm;" 
"To keep it alive;" "The old man likes them;" and so on. As a mat- 
ter of fact, on the production line, the ordinary director has had 
little time to learn his trade, so he gets the cast onto a set somehow, 
makes a close-up of each actor speaking his lines, and leaves a little 
room on the end of one for a fade-out. He has saved a half-day on the 
schedule, and the Front Office is pleased. But the picture generally 

There is, however, a very special reason for the use of close-ups. 
In a recent picture, the director had a big courtroom scene. The 
accused man was being questioned, the spectators were tense, the 
court reporters were busy, and the jury was listening. This was all 
shown in a fairly long shot. Then was shown a close-up of a girl in 
the very last row. She rises and speaks "Wait!"; back to the long 
shot as the whole courtroom turns to look at her. Again her close-up 
"I committed that crime!" Then the long shot spectators mill 
about lawyers are puzzled reporters run for phones and the 
judge raps for order. Now if this scene consisted of a close shot of 
the accused being questioned a close shot of the spectators a close 
shot of the jury listening a close shot of the court reporters working 
away a close shot of the girl in the back row rising she would then 
be merely one of a series of close shots and would not have an impor- 
tant spot in the film for the delivery of her speech. 

With regard to camera movements pans, dolly, zoom shots, 
etc., I do not know why most of these are used, as they can only slow 
up the action. For instance, the scene showing a character entering 
a room, crossing to the opposite side, when made with a dolly, has to 
follow the character every foot of the way, and is naturally quite 
dull. Without a dolly, the director would set the camera shooting 
at the door. The character would enter, close the door, and step 
out of the picture. Now the camera would be set up at the opposite 
side of the room, and the character would enter and go on with the 
scene. This would eliminate the long walk, and would speed up 
whatever action there was. There are cases where the long walk 
might add a dramatic note to the scene, but this is very seldom. 
Travelling shots are important when you want to see what is happen- 
ing to the stagecoach and the Indians. 

Matching the action between the long shot and the closer shot is 
a hangover from the old silent days. In those days, a director would 
have, for instance, a young man walk to the piano, begin to sit down, 


throw out his coat-tails from under him. While that was happening, 
the best cutters usually would cut to a closer angle, while the tails 
of the coat were still in mid-air. This was supposed to be the last 
word in smooth cutting. Nowadays, the best pictures are never 
made that way. A film cutter will complete whatever action he has 
to get over with one cut, then go on to the next cut, and finish that bit 
of action. He should never change a cut in the middle of a piece of 
business. It disturbs, rather than helps, the smoothness. 

Now comes editing. This is as different from cutting as newspaper 
reporting is from news editing. The news reporter can write three 
columns of words about a family being evicted from their home, and 
it might be very good writing. The editor can get the same story with 
all its values in half a column. 

So with the film editor. A director often spends months shooting 
a picture, and weeks cutting it. Then he previews it. In reel 2, the 
audience starts to yawn, and in reel 3 they are practically asleep. 
Then in the next reel, they suddenly pay attention to what's happen- 
ing, and actually stay awake for the rest of the picture. 

If such a situation happened in a news story, the editor would cut 
out the dull places drop the repetition, and fit the good parts 
smoothly together, keeping the story moving, interesting and crisp. 
That is the job of the film editor. The only trouble is that the job is 
so interesting that the whole executive force, from the President right 
on down, suddenly become film editors, and too many cooks spoil the 
broth. Some executives are beginning to recognize the importance 
of film editing. No one knows when they will get around to under- 
standing the sound-track. 

(At the end of the presentation a reel of film was projected demonstrating the technics 
of cutting for dramatic effect and shock, and cutting to rythm or tempo.) 



The editors present for convenient reference a list of articles dealing with subjects 
cognate to motion picture engineering published in a number of selected journals. 
Photostatic copies may be obtained from the Library of Congress, Washington, D. C., 
or from the New York Public Library, New York, N. Y. Micro copies of articles 
in magazines that are available may be obtained from the Bibliofilm Serivce, Depart- 
ment of Agriculture, Washington. D. C. 

American Cinematographer 

21 (December, 1940), No. 12 
Simplifying Makeup for Motion Pictures (pp. 549- 

570) W. D. FERGUSON 

British Journal of Photography 

87 (October 18, 1940), No. 4198 
Progress in Colour (pp. 507-508) 

87 (October 25, 1940), No. 4199 
Progress in Colour (pp. 515-517) 

87 (November 8, 1940), No. 4201 
Progress in Colour (pp. 539-540) 

Educational Screen 

19 (November, 1940), No. 9 
Motion Pictures Not for Theaters (pp. 379-381, 

402), Pt. 21 A. E. KROWS 

Electronics and Television and Short- Wave World 

13 (November, 1940), No. 153 
Columbia Colour Television, First Details of an 
Entirely New System (pp. 488-490) 

International Photographer 

12 (December, 1940), No. 11 
Process Technique (pp. 8-9, 23) 
pH Control (pp. 16-17, 19) D. K. ALLISON 

International Projectionist 

15 (October, 1940), No. 10 

Amplifier Power Output Data (pp. 7-8, 11) R. J. KOWALSKI 

An Instrument for Trouble-Shooting in Audio 

Amplifiers (pp. 13-14, 16-18, 24-25) C. E. MERVINE 




tf. S. M. P. E. 


22 (February, 1940), No. 2 
Der neue Askania-Zeitraffer (The New Askania 

Time-lapse Camera) (pp. 15-17) 
Der Beck-Lichtbogen in der Kinoprojektion (The 

Beck Arc for Motion Picture Projection) (pp. 

Neuzeitliche Aufnahme- und Widergabe-Objektive 

hoher Liestung (Recent High-Power Taking and 

Projecting Lenses) (pp. 19-20) 

Neue Schmalfilm-Gerate von Zeiss Ikon (New Zeiss 
Ikon Substandard Film Apparatus) (pp. 20-21) 

Der neue Schmalfilmzeitdehner mit Linsenscheibe 
des Institutes fur Kleinzeitforschung (New Sub- 
standard Film High-Speed Camera with Lens 
Disk from the Institute for High-Speed Investi- 
gation) (pp. 22-23) 

Kandem-Stufenlinsenaufheller mit Hochintensi- 
tats-Bogenlampe fur Filmbeleuchtung (Kandem 
Spot Light with High-Intensity Arc Lamp) 
(pp. 23-24) 

Neue Tongilmgerate (New Sound Film Apparatus) 
(p. 25) 

22 (March, 1940), No. 3 

Ein Spaltphotometer fur Netzanschluss in Differ- 
entialschaltung (A Slit Photometer for Connec- 
tion in a Differential Circuit) (pp. 31-33) 

Material und Arbeitsersparnis durch und bei 
Kombinationsaufnahmen (Economies in Time 
and Materials by and with Combination Expo- 
sures) (pp. 33-36) 

Die neue Koffer-Apparatur Bauer Sonolux II (New 
Portable Projector, the Bauer Sonolux II) (pp. 

Die Geyer-Film-Entwicklungsmaschine Nr. 187 der 
Geyer-Werke (The Geyer Film Developing Ma- 
chine, No. 187) (pp. 39-41) 

22 (May, 1940), No. 5 
Aus den Anfangen der Deutschen Kinotechnischen 

Gesellschaft (The Start of the German-Motion 

Picture Society) (pp. 59-63) 
20 Jahre Deutsche Kinotechnische Gesellschaft 

(Twenty Years of the German Motion Picture 

Society) (pp. 63-65) 





A. G. 





Jan., 1941] 



Entwicklungsrichtungen auf dem Gebiet der Ton- 
schriftarten (Tendencies in the Field of Sound 
Recording) (pp. 65-67) 

Ubersicht iiber den Stand des Problems; Vergleich 
zwischen Zacken- und Sprossenschrift (Synop- 
sis of the Problem: Comparison between Vari- 
able Area and Variable Density Recording) (pp. 

Der Beleuchtungsmesser Collux (Collux Exposure 
Meter) (pp. 70-72) 

22 (June, 1940), No. 6 

Die Grundlagen der Schnurschrift und neue Ver- 
fahren zu ihrer Herstellung (Fundamentals of 
Squeeze Track, and New Methods for Its Produc- 
tion) (pp. 76-81) 

Ein neues Prufverfahren fur die Herstellung von 
Zackenfilmkopien (New Test Method of Variable 
Area Duplicates) (pp. 8^-85) 

Kraftantreib fur Zeitdehneraufnahmen (Power 
Drive for High-Speed Exposures) (pp. 86-87) 
22 (July, 1940), No. 7 

Neuere Versuche mit Beck-Kohlen (New Investiga- 
tions with Beck Carbons) (pp. 91-95) 

Die Grundlagen der Schnurschrift und neue Verfah- 
ren zu ihrer Herstellung (Fundamentals of 
Squeeze Tracks, and New Methods for Its Pro- 
duction) (pp. 96-100) 

22 (August, 1940), No. 8 

Zuer Frage des gleichformigen Laufes von Tontra- 
gern (The Question of Uniform Running of the 
Sound Film Track) (pp. 108-112) 

Ueber den Lichtstrom der Ton-Optiken (Light 
Flux in Sound Optics) (pp. 112-115) 

Der Serien-Apparat von Ottomar Anschutz (Otto- 
mar Anschutz's Series Camera) (pp. 115-116) 
22 (September, 1940), No. 9 

Die Bildwandumgrenzung im Kinotheater (Screen 
Dimensions hi Motion Picture Theater) (pp. 

Uber Lampen mit nicht rotierenden Beckkohlen 
fur Stromstarken iiber 100 Ampere (Lamps with 
Non-Rotating Beck Carbons for Currents Ex- 
ceeding 100 Amperes) (pp. 126-128) 















Zur Frage des gleichformigen Laufes von Tontra- 
gen (Question of Uniform Running of Sound Film 
Track) (pp. 128-131) H. MULLER 

Motion Picture Herald (Better Theaters Section) 
141 (November 16, 1940), No. 7 
"Fantasia" Sound : Its Processes and Their Portent 

(pp. 7-8, 21) 

"Black Light" Grows as a Source of Theater Charm 
(pp. 11-13) 

Phillips Technical Review 

5 (April, 1940), No. 4 
Stereophonic Sound Reproduction (pp. 

107-114) K. DEBOER 

Photographische Industrie 

38 (April 24, 1940), No. 17 

Zur Helligkeitmessung in der Kinematographie und 
Projektionstechnik (Measurejent of Brightness in 
Motion Picture and Still Projection) (pp. 275-276) 





The 1941 Spring Convention will be held at Rochester, N. Y., with head 
quarters at the Sagamore Hotel. An especially interesting program of papers 
and presentations is being arranged by the Papers Committee, and Tentative 
Programs will be mailed to the membership of the Society about April 1st. In 
addition to the usual program of symposiums and technical papers, a special pres- 
entation and demonstration of stereophonic recording and reproduction is being 
arranged, to be held at the Eastman Theater in Rochester. Details will be an- 
nounced later. 

Early in March, hotel reservation cards will be mailed to the membership of the 
Society in order that accommodations at the Hotel may be reserved in advance. 
Members should not delay in returning these reservation cards. 


On December 18, 1940, the final meeting of the year was held at the Hotel 
Pennsylvania, New York. Three presentations of the Hollywood Convention, 
held last October, were re-presented here, namely: 

"The Twentieth Century Camera and Accessories;" D. B. Clark and G 
Laube, Twentieth Century-Fox Film Corp., Hollywood, Calif. 

"Production-Quality Sound with Single-System Portable Equipment;" D. Y 
Bradshaw, March of Time, New York, N. Y. 

"A Molded Plastic Screen with Contoured Surface;" R. O. Walker, Walker- 
American Corp., St. Louis, Mo. 

The first presentation was notable in respect to the fact that the authors had 
recorded their complete presentation upon 35-mm sound-film and shown in pic- 
tures all the details and operating mechanisms and functions of this outstanding 
development. The picture was introduced by Mr. Herbert Griffin. 

The second paper described the equipment used by March of Time in the pro- 
duction of the feature The Ramparts We Watch; and the third described a new 
type of screen surface molded of a flexible plastic material employing a specially 
contoured section about the holes provided for the passage of sound. 

Officers and Managers-Elect of the Section were introduced and a brief account 
was given by Mr. R. O. Strock, the Chairman-Elect, of the program contemplated 
for the coming season. 


At a meeting held on November 12th in the meeting rooms of the Western So- 
ciety of Engineers in Chicago, a paper was presented on the general subject of 



"Television" by Messrs. W. C. Eddy, Television Director, and A. H. Brolly, 
Chief Engineer of Balaban & Katz Corporation, Television Department, the engi- 
neers in charge of the erection of Television transmitter W9XBK. A demonstra- 
tion of equipment accompanied the presentation. The meeting was well attended 
and a lively discussion followed the presentation. 

Results of the election of officers and managers of the Section for 1941 were an- 
nounced, as follows : 

* J. A. Dubray, Chairman 

* I. Jacobsen, Secy.-Treas. 
** G. W. Colburn, Manager 

The other members of the Board of Managers, whose terms have not yet ex- 
pired, are: S. A. Lukes, Past- Chairman, and *C. H. Stone, Manager. 


One hundred and twenty-five members and guests of the Section attended the 
November 18th meeting held at the Review Room of Electrical Research Prod- 
ucts, Inc., Hollywood. Mr. J. H. Washburn, Chief Photographer of the Lockheed 
Aircraft Corp., presented a paper on "The Application of Motion Pictures to 
Aircraft Manufacture." The paper was illustrated with 16-mm pictures of test 
flights of finished planes. Much interest was shown in the subject, and consider- 
able discussion ensued. 

* Term expires December 31, 1941. 
** Term expires December 31, 1942. 




Volume XXXVI February, 1941 



American Standards and Their Place in the Motion Picture 
Industry J. W. McNAiR 113 

ASA Committee Proposes Method for Determining Speed of 

Film ' M. E. RUSSELL 119 

Operation of the Variable-Intensity Recording System ...... 


Ground-Noise Reduction Systems E. W. KELLOGG 137 

Some Laboratory Problems in Processing 16-Mm Black-and- 
White and Color-Films W. H. OFFENHAUSER, JR. 172 

Production-Quality Sound with Single-System Portable Equip- 
ment D. Y. BRADSHAW 180 

The Photographic Aspects of Television Operations 

H. R. Lubcke 185 

New Motion Picture Apparatus 

A New Treatment for the Prevention of Film Abrasion and 

Oil Mottle R. H. TALBOT 191 

Recent Developments in 8-Mm Copper-Coated High-Intensity 

Positive Carbons 


A Molded Plastic Screen with Contoured Surface 

R. O. WALKER 202 

Current Literature 207 

1941 Spring Convention at Rochester, N. Y., May 5th to 8th, 
Inclusive 209 

Society Announcements 213 





Board of Editors 
A. C. DOWNES, Chairman 




Subscription to non-members, $8.00 per annum ; to members, $5.00 per annum, 
included in their annual membership dues; single copies, $1.00. A discount 
on subscription or single copies of 15 per cent is allowed to accredited agencies. 
Order from the Society of Motion Picture Engineers, Inc., 20th and Northampton 
Sts., Easton, Pa., or Hotel Pennsylvania, New York, N. Y. 
Published monthly at Easton, Pa., by the Society of Motion Picture Engineers. 

Publication Office, 20th & Northampton Sts., Easton, Pa. 
General and Editorial Office, Hotel Pennsylvania, New York, N. Y. 

West Coast Office, Suite 226, Equitable Bldg., Hollywood, Calif. 
Entered as second class matter January 15, 1930, at the Post Office at Easton, 
Pa., under the Act of March 3, 1879. Copyrighted, 1941, by the Society of 
Motion Picture Engineers, Inc. 

Papers appearing in this Journal may be reprinted, abstracted, or abridged 
provided credit is given to the Journal of the Society of Motion Picture Engineers 
and to the author, or authors, of the papers in question. Exact reference as to 
the volume, number, and page of the Journal must be given. The Society is not 
responsible for statements made by authors. 


**President: EMERY HUSE, 6706 Santa Monica Blvd., Hollywood, Calif. 
** 'Past-President: E. ALLAN WILLIFORD, 30 East 42nd St., New York, N. Y. 
**Executive Vice-President: HERBERT GRIFFIN, 90 Gold St., New York, N. Y. 

*Engineering Vice-President: D. E. HYNDMAN, 350 Madison Ave., New York, 

N. Y. 
** Editorial Vice-President: ARTHUR C. DOWNES, Box 6087, Cleveland, Ohio. 

*Financial Vice-President: A. S. DICKINSON, 28 W. 44th St., New York, N. Y. 
** Convention Vice-President: W. C. KUNZMANN, Box 6087, Cleveland, Ohio. 

^Secretary: PAUL J. LARSEN, 44 Beverly Rd., Summit, N. J. 

* Treasurer: GEORGE FRIEDL, JR., 90 Gold St., New York, N. Y. 


**M. C. BATSEL, Front and Market Sts., Camden, N. J. 

*J. A. DUBRAY, 1801 Larchment Ave., Chicago, 111. 

*JOHN G. FRAYNE, 6601 Romaine St., Hollywood, Calif. 

*A. N. GOLDSMITH, 580 Fifth Ave., New York, N. Y. 

*ARTHUR C. HARDY, Massachusetts Institute of Technology, Cambridge, Mass. 
**L. L. RYDER, 5451 Marathon St., Hollywood, Calif. 

*T. E. SHEA, 195 Broadway, New York, N. Y. 

*R. D. STROCK, 35-11 35th St., Astoria, L. I., N. Y. 

Term expires December 31, 1941. 
**Term expires December 31, 1942. 


J. W. McNAIR** 

Summary. The American Standards Association is a federation of trade associa- 
tions, technical societies, and departments of the Federal Government. It was organ- 
ized in 1918 as a result of the country's experience during the World War, and has 
since served as the national clearing house for standards. Some 400 American Stand- 
ards have been approved to date in a wide variety of industrial fields and in the field of 
industrial and public safety. 

Some years ago at the request of the Society of Motion Picture Engineers, the Ameri- 
can standards Association organized a Committee on Motion Picture Standards. 
This might be said to be the beginning of national standardization in the photographic 
field. The committee brought together, under the sponsorship of the SMPE, the 
Academy of Motion Picture Arts and Sciences, the Acoustical Society of America, 
and a wide circle of scientific, engineering, and commercial groups interested in cine- 
matography. Some of the standards approved by the ASA through the cooperative 
uork of these many groups have become world-wide. There are 33 standards and 
recommended practices now before the ASA for approval as a result of long and 
arduous work by the motion picture Sectional Committee. 

The association is also very actively engaged in the development of national stand- 
ards for photography. A draft standard for determining photographic speeds of cer- 
tain types of negative materials will probably be published in a few weeks for trial and 

Six years ago at the request of the Society of Motion Picture Engi- 
neers the American Standards Association organized a committee on 
Motion Picture Standards. This committee brought together all 
the leading groups that were connected with the production of mo- 
tion pictures. There was the Research Council of the Academy of 
Motion Picture Arts and Sciences. There was the Acoustical Society 
of America, the Optical Society of America, the American Society of 
Cinematographers, the Motion Picture Producers and Distributors 
of America, in addition to a large number of the leading manufac- 
turing companies, trade associations, and a department of the Fed- 
eral Government. 

* Presented at the 1940 Fall Meeting at Hollywood, Calif., received October 
11, 1940. 

** American Standards Association, New York, N. Y. 


114 J. W. McNAiR [J. S. M. P. E. 

Prior to the time that this committee was formed there had been 
some standardization work on items relating to cameras, to film per- 
forations, to motion picture electrical devices, to projectors, etc. 
This early standardization work, as you know, was carried out largely 
by the Society of Motion Picture Engineers, and it was entirely in 
the field of silent pictures. 

With the advent of sound film many new problems arose. More 
societies and trade associations were formed, not only in this country 
but abroad. The Society of Motion Picture Engineers was no longer 
the only organization doing standardization work. The standardiza- 
tion activities of such groups as the Academy of Motion Picture Arts 
and Sciences were also becoming important, and there began to be a 
good deal of overlap in the work of various groups. Under the cir- 
cumstances it seemed advisable to combine the several sets of stand- 
ards then in use into one consistent set of American Standards. 

The American Standards Association is a federation of national 
trade associations, technical societies, and departments of the Federal 
Government. It was organized in 1918, as a result of the country's 
experience during the World War with the need for national indus- 
trial standards and for a uniform gauging practice. For twenty years 
industry and government groups have been using this agency to solve 
their technical problems. Through it some four hundred American 
Standards have been approved in a wide variety of industrial fields 
and in the field of industrial and public safety. These standards are 
today used throughout industry as well as by municipal, state, and 
Federal governments. Some six hundred organizations are taking 
part in the work through about three thousand representatives en- 
rolled on ASA technical committees. The standards so approved 
run all the way from performance standards for domestic gas ranges 
to specifications for Portland cement and sizes for pipe flanges and 

The American Standards Association differs materially from most 
organizations. First of all, being a federation of trade, technical, and 
governmental groups in all fields it does not represent any particular 
point of view. Of itself it carries on no testing activities, engages in 
no product research work. All this is done by the national groups 
and the companies who are members of the Association or who are 
taking part in the work. For instance, the American Petroleum In- 
stitute, the Association of American Railroads, the American Insti- 
tute of Electrical Engineers, the Army and Navy Departments, are 


all carrying on standardization activities as an important part of their 
own programs. The American Standards Association serves as a 
clearing-house for their activities and provides the machinery for the 
development of standards on an inter-industry basis. As an idea of 
how necessary this coordinating job can be, the power companies 
which make up the membership of the Edison Electric Institute pur- 
chase enormous quantities of insulated wires and cables every year. 
The railroads also use wire and cable, and so do the electrical con- 
tractors. If each of these groups develops separate sets of specifica- 
tions for wires and cables, none of the groups gets the advantage of 
economies resulting from one inter-industry set of specifications. 
A company like the Walworth Company, a large manufacturer of 
valves and fittings, sells to half a dozen industries. Before the develop- 
ment of American Standards for flanges and fittings this company 
had to do most of its work on special orders. It had no way of bring- 
ing its customers together to confer on materials and sizes, even 
though such action was to the customers' advantage. 

Unlike the standardizing bodies of most other countries, the Ameri- 
can Standards Association is not subsidized in any way by govern- 
ment. It is supported entirely through memberships. Sometimes 
the financial support has not come in fast enough to carry through 
the jobs brought to it by industry, for the work has constantly had a 
tendency to grow faster than the membership. The Association was, 
however, organized by industry to do industry's standardizing jobs; 
and so far industry has paid the way. The Association has a Consti- 
tution and By-Laws, and definite Rules of Procedure. Beyond that, 
it is run by the trade associations and other national groups that make 
up the primary membership and that are represented on the two gov- 
erning bodies of the Association. 

Without going too deeply into the details of how an American 
Standard is approved, a few words would be appropriate about Ameri- 
can Standards Association methods. One basic principle underlies 
all ASA work that any group having a substantial interest in the 
subject matter of a standard has a right to a voice in the develop- 
ment of that standard. The job of the Association is to bring together 
manufacturers, distributors, consumers, technical specialists, and 
others directly concerned with the work and to provide the ma- 
chinery whereby these groups, themselves, can arrive at mutually 
satisfactory solutions of their technical problems. 

The Association of itself initiates no work. It takes up a new proj- 

116 J. W. McNAiR [j. s. M. P. E. 

ect only upon request of a responsible organization or group. The 
project may deal with an existing standard already in general use, or 
it may involve the development of an entirely new standard. It may 
be any one of a wide variety of types dimensional standards ; speci- 
fications for materials; methods of test; performance specifications; 
methods of analysis; definitions of technical terms; industrial safety 
codes; industrial health codes ; building codes; etc. 

For flexibility, several methods are provided by which American 
Standards may be developed and approved. However, all are based 
on the democratic principle that every group having an interest in the 
project has a right to take part in its development, whether it holds 
membership in the organization or not. The term "American Stand- 
ard" means approved with the assent of every group having a sub- 
stantial interest in the finished job. The technical committees which 
work on these jobs are like miniature legislatures set up along indus- 
trial lines. It is not unusual for them to include representation of as 
many as forty to fifty national organizations. 

Naturally, the methods just described do not result in national 
standards overnight. The more groups involved and the more com- 
plicated the problems the longer it takes to get results. Some stand- 
ardization jobs have been completed in a few weeks. Some jobs take 
years. However, when a conclusion has been reached by the demo- 
cratic methods outlined above, it is pretty likely to be widely ac- 


With any fast-growing industry the problem of when to standardize 
and what to standardize becomes a very important one. If the de- 
velopment is allowed to proceed uncontrolled, confusion is likely to 
arise because of the diversity of practices, designs, and dimensions 
that come into use. If standardization is undertaken too soon, how- 
ever, or is unwisely carried out, this may also hamper the growth of 
the industry. 

The Society of Motion Picture Engineers, which was organized in 
1916, pioneered in the field of motion picture standardization. In 
fact, it was started with the avowed purpose of "standardization of 
mechanisms and practices" employed in the motion picture indus- 
try. Some of the early standards adopted by it are still in use today. 

The first attempt at national standardization was made in 1928, 
when the Society of Motion Picture Engineers submitted some stand- 


ards to the American Standards Association for approval as Ameri- 
can Standards. These related to professional motion picture film 
and equipment, exclusively. There were no competing standards 
in the field, and by common adoption these early SMPE standards 
have become practically international in scope. The situation with 
regard to 16-mm film and equipment, however, was quite different. 
As you know this equipment is used educationally and industrially 
and for almost every amateur use. It is in the standardization of 
16-mm sound-film and equipment that the value of the American 
Standards Association to the motion picture industry had its first 
real test. 

Most of you know the story of the 16-mm sound-film standards- 
how European industrial practice placed the sound-track to the right 
of the film, the exact reverse of American practice. American titles 
or printed matter when run through a foreign projector appeared as 
mirror images, and, of course, the sound part of the film could not be 
run at all. The chief film-producing countries were about equally 
divided as to use of the two practices. Commercial interests on both 
sides were losing money when the problem was thrown into the lap 
of the American Standards Association. That very day a cablegram 
went from the ASA to the International Standards Association, and 
the machinery for solving the conflict was put into motion. Eighteen 
months later we had a world standard for 16-mm sound-film and 
equipment. The certainty that any films produced could be used 
interchangeably on any apparatus opened a world-wide market for 
educational and industrial films, giving increased impetus to the al- 
ready thriving 16-mm sound-film industry. 

The work of the motion picture committee did not end with ap- 
proval of these early standards. Right now there are a number of 
standards and recommended practices before the ASA for approval. 
These deal with terminology, dimensions, etc., of film and equipment, 
and with various principles and practices in common use throughout 
the industry. 


A few words would be appropriate about standardization work in 
another field, but one so closely allied to motion pictures that it is 
bound to be of interest here. In 1936, the International Standards 
Association asked the American Standards Association to assume the 
secretariat for an international project in the field of photography. 

118 J. W. McNAiR 

This of course stimulated the development of national standards. 
A large and broadly representative committee of more then 50 mem- 
bers was formed to take charge of the work, and under it a number 
of sub-committees are pegging away at specific phases of the prob- 
lem. The Committee already has a number of standards in draft 
form. One for determining photographic speeds of certain types 
of negative materials will probably be published in a few weeks. 
This is an important standard to the motion picture as well as to the 
photographic industry. It includes a definition of the quality of ex- 
posing radiation, exposure time, concept of speed, speed criterion, 
speed number intervals, etc. It will be issued for a period of trial and 
criticism by the photographic industry before being approved as an 
American Standard. 

The scope of this photography project is to include every phase of 
picture taking except cinematography; for example, physical char- 
acteristics and dimensions of exposing equipment, sensitive materials, 
methods of developing and printing, methods of determining speeds 
of sensitive materials, etc. The four main objectives before the 
Committee are (1) to develop a system of nomenclature and termi- 
nology which will eliminate the present confusion caused by the use 
of the same words and phrases with different meanings; (2) to agree 
on dimensional standards to bring about better interchangeability; 
(3) to agree on uniform methods of expressing characteristics of sen- 
sitive materials; and (4) to define tests and methods of measurement 
which at present are not well known or not uniformly used. 

In all this photographic work the ASA is tying in with the stand- 
ardization activities of the Society of Motion Picture Engineers and 
with every other substantially organized group in the field of photog- 

The whole purpose of the American Standards Association is to 
bring to full realization the value of standardization work carried on 
by members and cooperating groups. The American Gas Association, 
for example, has vastly increased the acceptance of gas appliances 
through utilization of the ASA machinery. The Association of Ameri- 
can Railroads has saved money for its members through agreements 
worked out in ASA technical committees. In mechanical, electrical, 
mining, building, and other fields, industry has been turning more 
and more to standardization to save money and increase efficiency. 
Today national standards, in the development of which every inter- 
ested group has had a part, are being recognized as an important 
factor in our economy. 



The American Standards Association is now publishing a proposed 
method for determining photographic speed of roll film, film packs, 
and miniature camera films. The method was drawn up by the 
committee on photographic standardization (Z38), whose project was 
discussed in a previous issue. 1 The method is being published for trial 
and criticism for a period of approximately one year, at the end of 
which time it will be considered for adoption as an American Stand- 
ard. 2 If it is finally approved as an American Standard, it is ex- 
pected that it may be used as a basis for recommended exposures for 
picture-taking and for assigning speed numbers to films. 

The proposed method for measuring speed is built about the follow- 
ing concept: 

Photographic speed is to be considered as inversely proportional to the mini- 
mum camera exposure which a negative material must receive in order that an 
excellent print may be made therefrom. 

This concept implies that speed can be measured by making the 
best possible print from each of a series of negatives (negatives which 
differ only in the exposure they have received) and then deciding by 
observation the minimum negative exposure that will lead to an ex- 
cellent print. Such a concept of photographic speed seems so simple 
and straighforward that one might wonder why it had not been 
adopted years ago. The answer is that the making of picture nega- 
tives and prints to determine speed is a very unwieldy process. Be- 
sides the many difficulties in obtaining negatives properly exposed 
using a certain type of subject and lighting, a very large number of 
prints must be made and their quality judged by a large panel of 
judges making a multitude of observations. 

Numerous systems capable of laboratory control have been pro- 
posed for measuring speed during the history of photography and 

*Reprinted from Industrial Standardization, Nov., 1940, p. 277. 
**Chairman, Sub-Committee 2 on Sensitivity to Radiant Energy, of the ASA 
Sectional Committee on Standardization in the Field of Photography. 


120 M. E. RUSSELL [j. s. M. P. E. 

several of them have enjoyed considerable popularity. Each of 
these systems has been designed to give results quickly and simply 
and with good reproducibility. Unfortunately none of them gave re- 
sults which necessarily agreed with picture-taking practice. Often 
the discrepancy between the calculated results and those obtained by 
actual picture-taking was so serious that photographers used the 
term "practical speed" to indicate that in practice the speed would 
be found to differ significantly from that obtained by the laboratory 

Since the success of any method of determining photographic speed 
depends upon the ability to operate it with reasonable rapidity and 
with a high degree of reproducibility as well as agreeing with actual 
picture tests, it was necessary that a laboratory method be found 
from which the results would be the equivalent of picture-taking. 
Within recent years such a method has been evolved in the field of 
sensitometry, and it is this method which is used in the proposed 
standard. 3 It has been found to give excellent correlation with 
picture-taking practice and at the same time to be a satisfactory 
method for laboratory manipulation. 

The method consists of plotting a characteristic curve (that is, 
density vs. log exposure) of the material being tested for a particular 
set of exposing and developing conditions and then determining certain 
constants from the characteristic curve. In the making of a charac- 
teristic curve the photographic material is given a series of exposures 
on an instrument known as a sensitometer. In this instrument the 
spectral quality of the radiation and the time of exposure are made 
to match the average conditions in photographic practice as closely 
as possible. The exact shape of the curve is influenced by the type of 
exposure and development used, making it imperative that these 
conditions be exactly those for which the characteristic curve is 
intended to apply. 

After the sensitometric strip has been exposed and processed the 
blackness, or density, of each of the exposed areas is measured by a 
densitometer. A typical characteristic curve is shown in Fig. 1. 

In Fig. 1 an attempt has been made to illustrate how the sensito- 
metric method operates. At the top of the figure is a series of prints 
made from the negatives shown immediately below them. A series 
of exposures was given so that a range from badly underexposed 
to Very dense negatives was obtained. The best possible print was 
then made from each of the negatives. It will be noticed that as the 

Feb., 1941] 



negative exposure increases the resulting print quality increases 
rapidly to a high value and then remains substantially constant for 
further increase in negative exposure. The speed of the material is 
determined from the negative exposure required to yield the first 
excellent print. 




FIG. 1. Illustrating the use of the sensitometric criterion for determining 
photographic speeds. 

The characteristic curve for the negative material used is shown in 
the lower portion of the figure. The slope, or gradient, of the curve 
at any given point indicates the rate of growth of density with change 
in log exposure and thus shows the difference in photographic effects 
resulting from a given difference in scene brightness. It will be seen 
that the gradient produced by the photographic material differs with 
the exposure. 

122 M. E. RUSSELL [j. s. M. P. E. 

In the illustration the negative which produced the first excellent 
print was so exposed that it used the portion of the characteristic 
curve indicated between the log exposure values A and B. The 
average contrast of the negative is indicated by the slope of a line 
drawn from the lightest to the densest part of the negative, which 
in this case is the line CB. Since the characteristic curve does not 
have a straight line throughout the range of densities covered by the 
negative, the gradient varies from one part of the negative to another, 
the lowest value being in the shadow region of the picture as indicated 

It has been found, as a result of a comprehensive research, that if 
the gradient in the deepest shadow portion is 0.30 of the average 
gradient for the entire negative, the negative is capable of yielding 
an excellent print. If the gradient for the deepest shadow is less than 
0.30 of the average gradient of the negative, inferior prints will re- 
sult. Thus it is evident that the concept of speed, namely, that speed 
is related to the minimum exposure required to give a negative from 
which an excellent print can be made, is equivalent to the sensito- 
metric criterion that the gradient for the darkest portion of the sub- 
ject shall be 0.30 times the average gradient of the negative. 

This method of determining speed has been substantiated not only 
by the statistical fact that sensitometric results agree excellently 
with those found by actual picture- taking practice but by theoretical 
considerations as well. Since the quality of a photographic print 
depends upon the manner in which brightness differences have been 
recorded by the negative and the print materials, it is clear that any 
method for properly evaluating speed must be based upon the ability 
of the material to record brightness differences. The new method, in 
laying emphasis on the gradient characteristics of a material, is 
distinctly superior to the previously used methods for determining 

As mentioned before, the shape of the characteristic curve is de- 
pendent upon the exposing and processing conditions used. The 
light-source used is that adopted by the International Congress of 
Photography in 1928, and consists of an incandescent lamp filament 
operated at a color-temperature of 2360 K and screened by use of a 
specified liquid filter to give radiant energy of a quality closely 
approximating that of mean noon sunlight. 

The development of roll films and film packs is carried out in a 
metol-hydroquinone developer approximating that used in photo- 


finishing houses of the United States. The developer specified for 
the miniature camera films is a slower-acting developer and is widely 
used in the processing of miniature camera negatives. 

For testing purposes the developer agitation must be equivalent to 
that obtained by a hand-agitated Dewar flask fitted with a device for 
holding the exposed sensitometric strip. All materials are developed 
to a specified value of average gradient rather than for a fixed 

The present publication deals only with a method of determining 
speed of a specific sample of photographic material. As yet the details 
have not been worked out for applying the method to assigning a speed 
number to a product as a whole. The latter phase of the problem is 
being considered and no doubt some recommendation will soon be 

Another important phase of the speed problem is the determination 
of recommended exposures for normal picture-taking practice. It 
is clear that the present specifications are not quite adequate for this 
problem since they indicate the minimum exposure required for 
excellent results only for a specific piece of material when used under 
a specific set of handling conditions. Recommended exposures must 
take into account variations in processing, in film sensitivity, in the 
measurement or estimation of scene brightness, and similar variables 
which cause the user to obtain slightly different effective exposures 
from what he expects. The average consumer must use an exposure 
value which includes a margin of safety such that he is assured under 
all conditions of at least enough exposure to produce excellent results. 
Such information can be used either in the form of printed exposure 
guides or in connection with exposure meters. 

Thus far the committee has not had an opportunity to give serious 
consideration to the details of this problem of determining recom- 
mended "calculator numbers" or "meter setting values." It is 
clearly recognized, however, if the whole photographic industry is 
to derive maximum benefit from the standardization project, that 
specifications must be drawn as soon as possible to extend the method 
for measuring speed of a specific sample to the assigning of a speed 
number to a product as a whole and to the determination of recom- 
mended meter settings. 

Sub-Committee No. 2 on Sensitivity to Radiant Energy, which 
prepared the proposed standard on photographic speed of films for 
consideration by the ASA committee, has the following membership. 

124 M. E. RUSSELL 

The names of the individual members are followed by the organiza- 
tions they represent on the sectional committee. 

M. E. RUSSELL, Eastman Kodak Com- S. McK. GRAY, Electrical Testing 

pany, Chairman Laboratories 

PAUL ARNOLD, Agfa Ansco Division of ALBERT F. HOGLE, Master Photo 

General Aniline and Film Corpora- Finishers of America; The Photo 

tion Finishing Institute 

WALTER CLARK, American Committee F. K. McCuNE, National Electrical 

of the International Congress of Manufacturers Association 

Photography, and Eastman Kodak BRIAN O'BRIEN, Optical Society of 

Company America 

RAYMOND DAVIS, National Bureau of ROWLAND S. POTTER, Defender Photo 

Standards, U. S. Department of Supply Company, Inc. 

Commerce E. D. TILLYER, American Optical 

HANS DESSAUER, The Haloid Company Company 

W. N. GOODWIN, JR., National Elec- D. R. WHITE, duPont Film Manufac- 

trical Manufacturers Association turing Corporation 

Other sub -committees are now at work on the program of the ASA 
Committee on Standardization in the Field of Photography (Z38) 
as follows : 

Physical dimensions of sensitive materials, specifically of unexposed, unprocessed 
sensitive materials and holders therefor (Sub-Committee 1} 

Supports for sensitive coatings (Sub- Committee 3} 

Exposing equipment: cameras, lenses, shutters, etc. (Sub-Committee 4) 

Photographic characteristics of illuminants (Sub- Committee 5) 

Processing equipment (Sub- Committee 6} 

Printing and projection equipment (Sub-Committee 7) 

Processing (Sub-Committee 8) 

Definitions; abbreviations and symbols; form and arrangement of published 
standards; numbering of standards (Sub-Committee 9} 

The Optical Society of America has the administrative leadership 
for the work of the ASA committee, with Loyd A. Jones as chairman, 
and J. W. McNair of the American Standards Association as secre- 


1 McNAiR, J. W.: "Progress Made in Photographic Standardization," Indus- 
trial Standardization, X (Dec., 1939), p. 293. 

2 These proposed standard specifications will be printed in the J. Opt. Soc. 
Amer. (Jan., 1941). 

3 JONES, L. A.: "The Evaluation of Negative Film Speeds in Terms of Print 
Quality," /. Franklin Inst., 227 (1939), p. 497. 

JONES, L. A., AND NELSON, C. N. : "A Study of Various Sensitometric Criteria 
of Negative Film Speeds," /. Opt. Soc. Amer., XXX (Mar., 1940), p. 93. 



Summary. A description of the system is given with line drawings illustrating 
the optical design together with operational characteristics of the system and an 
account of how it has been adapted to existing studio technic. 

The stage channels, which are all mobile, are described with illustrations, and 
methods of setting the channels up for operations are discussed. 

The RCA variable-intensity recording system was put into use at 
Twentieth Century-Fox Studios in November, 1938. Standard 
Class A track, 100 mils in width, is used for all dialog and release 
negatives and Class A push-pull single-width track is used for music 
scoring original negatives. 

Fig. 1 is a dragram of the optical system and shows the positions 
of the lamp A, galvanometer F, and the penumbra mask T. The 
lens B is a condenser lens and in combination with lens E forms an 
image of the lamp filament A on the galvanometer mirror. A por- 
tion of the recording light-beam at the slit is reflected by the lens 
mirror L back through the monitor system onto the monitor card. 
The lens G just back of the slit H forms an image of the galvanometer 
mirror in the objective lens / which images the slit H onto the film 
at an optical reduction that gives a recording light-beam 1 / 2 mil in 
height and any length up to 100 mils. V is the noise-reduction motor 
that actuates the moving penumbra mask U which is located a few 
thousandths of an inch back of the fixed penumbra mask. W is a 
selective mirror which diverts light from the recording light -beam 
downward through the cylindrical lens X onto the second mirror Y 
out through the splitting lens combination Z to the photocell Q. 
The system is designed with the splitting lenses so that it may be 
used with push-pull systems. The desired noise-reduction is ob- 

* Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received November 
4, 1940. 

** Twentieth Century-Fox Film Corp., Hollywood, Calif. 
t RCA Manufacturing Co., Hollywood, Calif. 



tained by mechanically adjusting the movable penumbra mask into 
the light-beam until the intensity of the light at the slit is reduced to 
its correct value. It is pulled down by the noise-reduction current a 
distance proportional in amplitude to the signal plus the desired 
margin. 1 

Fig. 2 is a photograph of the optical system with the photocell 
cover removed, showing the relative positions of the cell and splitting 

In installing the system in the studio it was desirable that as few 
operating changes as possible be made so as to reduce confusion. 
The motor system and operational features were kept as nearly stand- 


FIG. 1. Optical system. 

ard as possible with existing studio equipment. Sufficient tests were 
run to determine the exposure range and the linear modulation range 
of the system in order to adapt it to the existing processing conditions 
of the laboratory. 

The system has a linear exposure from no light to full intensity, 
the upper limit being determined by the intensity of the source, 
which in turn is limited by the maximum safe lamp current. Fig. 3 
is a curve showing the linearity of exposure light vs. deflection of the 
galvanometer from zero (no light) to 100 per cent modulation of the 
system (full light). 

Fig. 4 is a curve showing the input vs. output from the system and 
the distortion present. 


This system, making use of a constant- width aperture, has distinct 
advantages over the variable aperture in that distortion from film 
travel, and other distortions commonly associated with a variable 
aperture are considerably reduced. Also, the const ant- width slit 
permits better control of frequency equalization of the recording 
channel as there is no variation in film response with amplitude of 
signal at the higher frequencies. 

Fig. 5, curve A, is the electrical response of the system. Curve B 
is the response of the system through the PEC monitor, curve C is 
the overall recording characteristic including the galvanometer, and 
curve D is the response from the film measured through a sound-head 
which has been equalized for flat response from the SMPE test reel. 

FIG. 2. Optical system with photocell cover removed. 

To utilize more fully the advantages offered by this system it was 
felt advisable to start from a point sufficiently low on the penumbra 
curve in order that the system could accommodate a higher signal 

Fig. 6 is a typical track gamma curve, density vs. galvanometer or 
penumbra deflection developed to a gamma of 0.40. Hence for 100 
per cent deflection of the penumbra the mean density is approxi- 
mately 0.7, represented by point A. This represents the center- 
point of the penumbra where the moving vane passes behind the 
fixed vane. Modulation about this point therefore represents maxi- 
mum with the light modulating from zero to maximum and the den- 
sity modulating from its lowest value up to 0.83. 


C. W. FAULKNER AND C. N. BATSEL [j. s. M. P. E. 


































2O 40 60 &O 100% 


FIG. 3. Exposure light vs. deflection of galvanometer from zero 
to 100 per cent modulation. 


FIG. 4. Curve of input vs. output of the system, 

Feb., 1941] 




FIG. 5. Electrical response. 



: GAV 

MA 51 





























9 n 3 is IT 19 a 

FIG. 6. Typical track gamma curve. 


C. W. FAULKNER AND C. N. BATSEL [j. s. M. P. E. 

In order to conform to present laboratory practice, it was deemed 
desirable to keep the unbiased negative density at 0.5. This density 
is obtained at 17 per cent of the penumbra range and is represented 
at point B on the curve. The noise-reduction is taken at a point C, 
which is 10 db below B, and is obtained at 5.4 per cent of the pe- 
numbra range. 

In practice the moving penumbra vane is mechanically set at 
point C for no modulation. The application of sufficient signal 
carries the vane up to point B, which represents 100 per cent record- 
ing level or a modulation range from 0.3 to 0. 65 in density. Signals 

FIG. 7. Curve of strip printed from sample of Fig. 6. 

that carry the vane up to 0.7 represent peaks and unusually loud 
sounds which produce a maximum density of 0.83 on the negative. 
The negative range A to B is utilized in this manner for peak pro- 

A printed strip of Fig. 6 is shown on Fig. 7. This shows a print 
linearity up to 80 per cent of the penumbra range. The relationship 
of the points on the track gamma strip correspond to similar points 
on the print transmission curve. The region C' to D r represents the 
100 per cent recording range on the negative. D' to E' represents 
the portion of the negative from D to E and is reserved for overloads. 

All stage recording is done by mobile units. The trucks are two- 
ton Fords with special bodies designed to meet the requirements of 

Feb., 1941) 



the studio (Fig. 8). The recording equipment is mounted in a com- 
partment occupying the center portion of the truck body (Fig. 9). 
The rear portion of the body contains all the power equipment, selsyn 
master motor and storage space for the stage mixer and cables. The 

FIG. 8. (Upper) Exterior of mobile unit. 
FIG. 9. (Lower) Recording equipment in mobile unit. 

units are all a-c operated with standby A and B battery complement. 
They carry no power supply for the motor system, use being made of 
220- volt mains wherever they are available. In the event that loca- 
tion work necessitates auxiliary power, a Dodge 3 /4-ton truck is used. 
In this truck are mounted (Fig. 10) batteries and a motor-driven al- 
ternator of sufficient capacity to furnish 220-volt 3-phase power to 
operate the recording truck. A charging generator driven by the 



truck engine may be used either to float the batteries or charge them 
during idle periods, thereby avoiding the necessity of recharging from 
power mains. 

FIG. 10. (Upper) Power supply in mobile unit. 
FIG. 11. (Lower} Re-recording and scoring amplifier installation, 

The audio channel equipment comprises a four-position mixer con- 
taining four two-stage microphone amplifiers which, with the micro- 
phone and boom equipment, constitutes the equipment normally 


used on the set proper. Transmission from the mixer is through a 
90-HI-78 high-pass and an 8000-cycle low-pass filter into the main 
gain amplifier which drives the modulator. The fixed dialog equali- 
zation is located in this amplifier. The noise-reduction amplifier is 
bridged off the bridging bus through a 5000 to 500-ohm pad. The 
monitoring amplifier is bridged across the main amplifier interstage 
for direct monitoring. The volume indicator is also bridged across 
the main amplifier output. 

The output of the PEC monitoring photocell is transformed and 
fed into the PEC voltage amplifier. Provision is made for auto- 
matically switching the output of this amplifier to the input of the 
monitoring amplifier during takes so that both truck and mixer moni- 
toring is normally PEC. The mixer may listen to direct monitoring 
during takes by switching to the bridging bus by means of an AB key 
at the console. 

The recorders, which are of the magnetically driven drum type, 
are equipped with the visual monitoring card which is of considerable 
aid to the operators in checking levels, setting up, and observing the 
operation of the modulator system in general. Operating experience, 
however, has proved the desirability of having a reliable means of 
actually measuring the transmission of light through the recording 
slit. It is particularly important that the operator be able to check 
the biased and unbiased position of the shutter in respect to the mid- 
point position of the galvanometer. To accomplish this the recorders 
are now being equipped with a light-meter. 2 

This device is based on the principle of applying a fixed alternating 
voltage to the d-c polarizing voltage of the photocell. The output of 
the photocell is amplified through the PEC amplifier system whose 
output is read in terms of db on an ordinary power-level indicating 
meter. A variable attenuator between the amplifier and the meter 
is provided with steps that represent the output from the cell with 
100 per cent light, 50 per cent light, and from this point on down there 
are twenty steps of one db each. This arrangement permits the 
operator quickly to check his 100-per cent and 50-per cent illumina- 

The final noise-reduction shutter position is then obtained by ad- 
justing the zero shutter current until the output meter reads the re- 
quired level in db below the 50-per cent point. At present, with the 
use of 10-db reverse bias, this setting is 20 db below 50 per cent. 

Synchronizing marks and sound-track slates are photographed on 


the film in the sound-track area while the machine is at rest. This 
procedure is possible since all motors are selsyn interlock. 

The re-recording and scoring amplifier installation is shown in 
Fig. 11. All amplifier equipment is the same as that used on the 
trucks. It is arranged in the rack, however, best to serve the needs 
of the plant and will accommodate simultaneously scoring and re- 
recording operations. The scoring mixing console is arranged so 
that, by patching, the two four-position mixers may be fed into sepa- 
rate channels for dual recordings. 

The re-recording channel reproducing complement consists of ten 
film reproducers, nine of which are dummies employing the RCA 
rotary stabilizers and one is a magnetically driven drum film-phono- 
graph. They are all equipped for push-pull or standard reproduction. 
The PEC monitoring is supplied as in the mobile units and may be 
switched from direct to PEC at the mixing console as desired. 

The recorders are the same as the mobile units with the exception 
of the push-pull music scoring machine. The optics of this machine 
are different from those of the standard RCA variable-intensity in 
that a double penumbra is formed at the slit. These operate 180 
degrees out of phase and are separated by a septum so that the two 
tracks do not overlap. 


1 DIMMICK, G. L. : "The RCA Recording System and Its Adaptation to 
Various Types of Sound-Track," J. Soc. Mot. Pict. Eng. f XXIX (Sept., 1937), p. 

DIMMICK, G. L.: "Optical Control of Wave-Shape and Amplitude Character- 
istics in Variable-Density Recording," J. Soc. Mot. Pict. Eng., XXXIII (Dec., 
1939), p. 650. 

2 DAILY, C. R. : "Use of an A-C Polarized Photoelectric Cell for Light-Valve 
Bias Current Determination," J. Soc. Mot. Pict. Eng., XXXIII (Oct., 1939), p. 394. 


MR. LORANCB : What are the slit image dimensions on the film ? 

MR. BATSEL : l / 2 mil high by 100 long. 

MR. LORANCB : What is the lamp current, and what type of lamp is necessary? 

MR. BATSEL: We use a bayonet base T-8 bulb curved filament recording 
lamp rated at 10 volts, 7.8 amperes. The lamp current is at present about 7 

MR. LORANCE : That is on the usual standard negative emulsions? 

MR. BATSEL: Yes, the system is operated so that 17 per cent of the total ex- 
posure produces the unbiased negative density. 

DR. FRAYNE : Is it possible to expose 1301 emulsion with this type of optic? 


MR. BATSEL: We can expose 1301 stock. If developed at present gammas 
it would require about 0.4 ampere increase in lamp current, which would still 
leave a safe margin with the 7.8-ampere lamp ; or we could expose the unbiased 
negative density at a higher point than 17 per cent of the exposure range and 
work with lower lamp current. 

MR. RYDER : Is sufficient exposure available for the proper exposing of either 
the 222 or the 1302 fine-grain film? 

MR. DIMMICK: If a high-pressure mercury -vapor lamp is employed as a 
light-source, fine-grain films may be exposed without difficulty at normal nega- 
tive gamma. If the negative gamma is increased, it is possible to expose fine- 
grain films with an incandescent lamp. In the latter case, an optical compensator 
is placed in the recording optical system to reduce the effective overall gamma to 

It also has the advantage that the full characteristics of the film will be com- 
pensated, at the same time giving greater depth to the recording motor. 

DR. FRAYNE: I am very much interested in Mr. Dimmick's remarks about 
the possibility of anticipating the film distortion by putting in a counteracting 
distortion. We have tried such a scheme but failed because we never could tell 
in advance exactly what the film characteristics would be. They change from 
time to time, and from batch to batch of emulsion. 

MR. DIMMICK: The present low-gamma variable-density system is not 
linear for high print transmissions. In addition, there are variations in the 
overall transmission curve due to normal variations in the negative and print 
developers. In my opinion, it is better to make the average transmission char- 
acteristic linear and accept normal variations from this condition, than it is to 
accept both the non-linearity and the variations. 

MR. RYDER: In other words, if we have a known distortion in a given direc- 
tion and compensate largely for that distortion, and then take the average around 
that point, we would be better off than taking the fixed amount of distortion. 


MR. ALBIN : Mr. Dimmick's point is good theoretically, but it does not work 
out hi practice. If we know the average variation in film distortion and make an 
average correction, then part of the time there is over-correction and the dis- 
tortion becomes tremendous. It is apparently far better to be under-corrected 
than over-corrected. 

In other words, considerable compression of the type introduced by the film 
characteristic may be tolerated by the listener, but practically no expansion is 
permissible. The plan of pre-distortion, anticipating film distortion, is very ad- 
mirable provided that film compression is not less than predicted and expansion 

MR. DIMMICK: Perhaps Mr. Albin would prefer a degree of compensation 
that would make the transmission characteristic linear at one limit of the proc- 
essing tolerance and allow it to be slightly curved in its present direction for the 
other limit. The overall result would still be an improvement over present prac- 

MR. LORANCE : Mr. Batsel said that he used alternating current on the photo- 
cell in the exposure meter. Over what range is the cell linear? Are there any 
troubles with non-linearity? 


MR. BATSEL : We are able to work the meter through a linear range of 36 to 40 
db, which is sufficient for this application. 

MR. LIVADARY: If I understand Mr. Dimmick correctly, the pre-distortion 
that may be put into the optical system of the intensity recorder is to be such as to 
compensate for future distortion introduced by the characteristics of the proc- 
essed film, and thereby result in a linear output from the film within the modula- 
tion range of the apparatus. 

In view of the strong argument that such a linear record is not desirable, as 
exemplified by efforts made to compress variable-area records that are inherently 
linear, does that mean that linear reproduction is satisfactory in variable-intensity 
recording, whereas it fails in the variable-area case? 

MR. DIMMICK: Most of us agree that if compression is used, it should be of the 
type that does not produce wave-shape distortion. It would therefore seem ad- 
visable first to make the variable-density system linear and to apply non-distort- 
ing electronic compression to any extent desired. 

There seems to be some misunderstanding as to how the compensation is 
introduced. A correcting plate can be placed in the light-beam of the 
system, the plate being of such shape as to make the relation of intensity to 
deflection logarithmic. In other words, the recording system can be given a 
gamma different from unity. 

DR. DAILY: In determining optimal film-processing conditions, have you 
observed that distortion calculated from transmission measurements of printed- 
through 116 strips do not always provide a reliable index of the sound quality of the 

MR. BATSEL: We prefer to use print-through strips from actual track gamma 
strips made on the recorder. The negatives do not show the usual shoulder ob- 
served on the H&D strips; consequently actual transmission measurements of 
the prints are much more indicative of recording results. 

MR. ALBIN : In the curve showing the relation between negative exposure and 
positive transmission, which was quite linear over the usable range, how was the 
transmission of the positive measured with a densitometer, measuring diffuse 
density; or by sound-head, measuring projected density, for example? 

MR. BATSEL: Measurements are made in a standard sound-head reproducer. 
Transmission values are read by means of the RCA ultrasensitive d-c meter from 
the output of the sound-head photocell. 


Summary. This paper is not a discussion of any specific commercial ground- 
noise reduction system, but rather of general principles, and is an effort to formulate 
a statement of the desired characteristics of a ground-noise reduction system, in terms 
of such factors as promptness of opening, peak reading, and filtering. In this it is 
assumed that anticipation is not employed. It is desirable to limit the filtering to a 
single stage of resistance-capacity filtering (or equivalent) . Slow closing helps filter- 
ing and peak reading. The better the peak-reading properties of the circuit and the 
less the filtering delay, the smaller can the margins be made without causing too fre- 
quent clipping. 

A number of circuits are discussed which have been proposed for improving the 
filtering without sacrificing quickness of opening or reasonably rapid closing. 

In some operations, anticipation is entirely practicable, and if this is done, it 
appears possible to provide an almost perfect envelope current. 

The pioneers of photographic sound recording had to solve many 
problems, some of which are still with us. One such chronic problem 
is ground-noise, due to scratches and dirt on the film. L. T. Robin- 
son 1 realizing that most of the noise came from specks in the clear 
area of the film rather than from holes in the black areas, proposed to 
bias the galvanometer during periods of low modulation, so that the 
clear area was just enough to accommodate the modulation. He em- 
ployed the simple circuit shown in Fig. 1. Recording at that time was 
being performed by galvanometers of the oscillograph type which had 
a resistance of about two ohms. In such a low-impedance circuit, it 
is a simple matter to employ inductances and resistances to separate 
the high and low-frequency components, but difficult to take ad- 
vantage of condensers. The circuit used by Robinson was by no 
means perfect, but it demonstrated the important principle and did 
reduce the ground-noise. 

C. W. Hewlett 2 and C. R. Hanna, 3 who were also among the first 
to contribute to the solution of this problem, performed their recti- 

*Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received December 
3, 1940. 

**RCA Manufacturing Co., Camden, N. J. 




LT. S. M. P. E. 

fying functions in the amplifier, thereby facilitating the filtering and 
making it possible to eliminate from the waves being recorded any 
distortion due to loading the circuit by the rectifiers. Figs. 2 and 3 
show in simplified form the Hewlett and Hanna circuits. 

In variable-width recordings of the unilateral type used prior to 
1932, biasing the galvanometer puts the low modulation recording at 
the extreme edge of the sound-track as shown in Fig. 4. There is an 
objection to doing this, because the recording optical systems gave 


FIG. 1. Original ground-noise reduction circuit of L. T. Robinson. 
FIG. 2. Circuit of C. W. Hewlett. 
FIG. 3. Circuit of C. R. Hanna. 

slightly poorer resolution at the edges of the track, and the reproduc- 
ing optical systems, in many cases, gave poor illumination of the 
edges of the track. There was also danger of entirely losing some of 
the sound in case of sidewise weaving of the film. H. McDowell 4 - 5 
corrected this fault by employing a shutter to intercept a part of the 
recording light-beam, instead of biasing the galvanometer. This left 
the small modulation at the middle of the track, the unused track 
area being unexposed and clear in the negative and therefore black in 
the positive, as illustrated in Fig. 5. The work of McDowell, supple- 
mented by that of others, 5 and the successful application of the 


biasing system to variable-density recording 6 directed the attention 
of the industry to the advantages to be gained from ground-noise 
reduction systems as applied to variable-area recording. Commercial 
designs of shutters 7 and suitable amplifiers 8 were made available. 

The adoption of the symmetrical track 9 by RCA obviated the 
original objection to biasing the galvanometer, for in the case of the 
symmetrical track, biasing the galvanometer leaves the low modula- 
tion at the middle of the track (Fig. 6). Other considerations, how- 
ever, resulted in a return to the use of shutters, and by suitable modi- 
fications a system employing shutters can be readily adapted to 
various types of recording such as push-pull track and the penumbra 
system of variable-density recording. 11 - 12 

There have been numerous variations in the circuits for providing 
the control current, generally representing slightly different points of 
view of those who are designing equipment or are using it, or a differ- 
ent emphasis on the importance of several factors. It is the purpose 
of this paper to review the arguments for and against a number of 
the actual or proposed circuits. 


In general terms, it may be said that the purpose of a ground-noise 
reduction system is to keep the sound-track light transmission at the 
lowest average value that will permit the required modulation, since 
any additional (unmodulated) light simply increases the noise. The 
changes in average transmission, however, must not be rapid enough 
to produce audible sounds. These statements apply equally to vari- 
able-density and variable-width systems. The absolute magnitude of 
the light modulation should not be altered by anything that the 
ground-noise reduction system does. Thus when it reduces the 
average light reaching the photocell, this light is modulated by a 
higher percentage by the audio waves than would otherwise be the 
case. In fact, we might say that the ground-noise reduction system 
so regulates the average light in relation to the magnitude of the 
recorded waves that the modulation is kept as near 100 per cent as is 
practicable, without running too much risk of its actually reaching 
100 per cent. Since the fundamentals of ground-noise reduction as 
outlined above, are essentially the same in both the density and the 
area systems, we can discuss the ground-noise reduction requirements 
in terms of its application to a variable-area track, and assume that 
the latter is one of the type made with a shutter, such as illustrated 



[J. S. M. P. E. 

in Figs. 5 and 7. The axis of the sound-wave trace produced by the 
galvanometer vibrations is a straight line parallel to the axis of the 
track, and a separate trace is produced of the movements of the 
shutter. When the modulation is low, the shutter advances so that 
a very minute light-spot strikes the film, producing a narrow black 
line on the negative, or a narrow clear line on the print. As the modu- 
lation increases, the shutter backs away, and should do so in such a 
manner that it at no time prevents the recording of the complete 
audio waves. This condition is not entirely achieved in practice, 

m JJ Ad* u 

FIG. 4. Track produced by galvanometer bias; 


FIG. 5. Track produced by shutter; McDowell. 

FIG. 6. Symmetrical track with biased galvanometer. 

FIG. 7. Symmetrical track made with shutter. 

and we refer to the system as "clipping" the tops of the waves. If 
there is not a great deal of clipping, it may be difficult to detect the 
effect in listening tests, but there can scarcely be any doubt that it is 
detrimental, and one of the principal purposes of our efforts to im- 
prove ground-noise systems is to minimize the clipping. On the other 
hand, we do not want the shutter to back away to an unnecessary 
degree, for this leaves large clear areas on the print, in which dust and 
film abrasions will produce noise. 

The ideal characteristic of a ground-noise reduction system has 
been aptly described by saying that the shutter trace should follow 


the "envelope" of the sound-waves. If we were given an outline of 
audio waves such as an enlarged recording or an oscillograph trace, 
there is little doubt that a large number of engineers, if asked to draw 
an envelope, would draw substantially the same curve, but if we 
attempt to define an envelope, we find that there are cases of am- 
biguity. We speak of the envelope trace as following along and just 
touching the tops of the waves, but what if some waves are higher 
than others? If there are two tall waves and a short one between, 
we should, in general, say that the envelope curve will ignore the 
shorter wave and touch only the higher peaks, but this will depend 
upon how far apart the high peaks are. Fig. 8 shows two possible 
envelope curves for the same audio waves. Obviously if the modula- 
tion drops, the envelope curve should follow down. It is not difficult 
to see when it should follow down and when not, provided we can look 


FIG. 8. Two types of envelope curve. 
FIG. 9. Delayed discharge and expo- 
nential discharge curves. 

at the complete recording, and see what is ahead. But if we do not 
know what is coming, it is not so simple a matter to decide whether 
a given high peak which we have just passed is the last one of a series 
and it is now time to begin to drop ; or how rapidly can we drop with- 
out cutting the next peak (which will presumably be of lower ampli- 
tude, but whose exact amplitude we do not know). We should like 
to require of our ground-noise reduction amplifiers a degree of in- 
telligence and prognostication that we do not ourselves possess. 
Realizing that the drawing of a logical envelope wave requires that 
we know what is coming, we can see that our system can be made 
much more satisfactory if anticipation is possible; or, in other words, 
if the waves can reach the ground-noise reduction system slightly in 
advance of their actual recording on film. Anticipation through the 
employment of a delay circuit is mentioned by Silent and Frayne, 6 

142 E. W. KELLOGG tf. s. M. P. E. 

and others. Anticipation can also be more simply provided in cer- 
tain cases, such, for example, as in re-recording, or in making direct 
positives as described in a paper by G. L. Dimmick 13 ** and in 
several U. S. patents of earlier date. 14 - 15 Let us, however, for the 
moment confine ourselves to the case where anticipation is not 
possible, and ask ourselves what we should like our ground-noise 
system to do. 

After a period of little or no modulation, the arrival of any audio 
wave, in general, means that more waves will follow, and probably in 
increasing amplitude. The first wave will inevitably be clipped un- 
less it is smaller than the shutter opening at the time. The shutter, 
however, should open as promptly as possible and thereby minimize 
the number of clipped waves. As soon as modulation of appreciable 
amplitude is present, it is permissible to increase the width of the clear 
film, for it is well recognized that more ground-noise can be tolerated 
when masked by modulation. Therefore, as soon as there is any 
evidence of audio waves to be recorded, the shutter should open far 
enough at least to increase the clearance to something substantially 
greater than the narrow zero-modulation clearance. Thereafter, the 
shutter should continue to open at a rate depending upon the actual 
magnitude of the incoming audio waves. 

During a sustained train of waves, the shutter should, of course, 
remain open by an appropriate amount. When the waves begin to 
decrease in amplitude, the shutter may begin to close. It is necessary 
that the shutter close with sufficient rapidity so that we do not hear 
conspicuous ground-noise after the modulation has ceased or reached 
a low value. There are two factors which contribute to making 
moderately slow closing permissible. Reverberation both in the 
recorded sound and in the reproducing auditorium prevents an abrupt 
stop of the recorded sound, and thus masks ground-noise. There is, 
perhaps, also an accommodation effect by reason of which the ear is 
less sensitive to faint sounds immediately after a relatively loud 
sound. 16 * It is desirable to take advantage of these factors, and not 
close the shutter any more rapidly than is required to keep the ground- 

*Stevens and Davis 16 (Hearing, p. 217) discuss this under the headings "Fatigue" 
and "Persistance of Sensation." Experimenters do not find anything comparable 
with adaptation of the eye to light-intensity. Tests by B6k6sy are reported 
showing that the ear does not recognize any difference between an 800-cycle tone 
which ends abruptly and one which decays 60 db in 0.1 second. 


noise from being noticeable. This will in turn depend somewhat upon 
the magnitude and type of the ground-noise. Thus the steady hiss, 
characteristic of density recording, appears to require quicker closing 
than the more haphazard noises which result from too much clear 
film in a variable-width track. 

There are two distinct advantages in slow closing. One is that it 
greatly reduces the ripple which has to be removed by a stage of 
filtering; the second is that every time the shutter closes there is 
danger of clipping waves the next time it has to open. Thus, it would 
be unfortunate for the shutter to close if it would have to open again 
immediately afterward for new high modulation. If the shutter starts 
to close immediately after the passage of a peak in the audio waves, 
it will attempt to follow the individual audio waves, especially if 
these are of comparatively low frequency. Such movements of the 
shutter are obviously not a part of the desired envelope character- 
istic, which should follow straight across from top to top of the waves. 
It would appear logical, then, to ask that the shutter remain sub- 
stantially at the full opening called for by the magnitude of the last 
peak, for a period corresponding to one cycle of the fundamental 
audio wave. If after this interval another wave of equal magnitude 
fails to arrive, the shutter may begin cautiously to close. It would 
call for unjustifiable complication to specify that our shutter should 
adopt a different period of expectant waiting, depending on the 
frequency of the tone being recorded. We shall therefore specify 
a substantially fixed time which will cover the fundamental period 
of most audio tones to be recorded. Since voice is the most important 
of the audio tones and only exceptionally do men's voices drop much 
below 100 cycles, it might be in order to specify a waiting period of 
the order of ten milliseconds as desirable. Thus a closing character- 
istic after modulation has stopped, as illustrated by curve 1 of Fig. 9, 
would be desirable rather than the one indicated by curve 2 of the 
same figure. It is, of course, not important as a practical matter 
that conditions remain absolutely static for the suggested ten milli- 
seconds, and curve 1 has been drawn to represent a recovery or closing 
that is not perfectly flat, but relatively slow for the first ten milli- 
seconds. The filter, which is a part of every ground-noise reduction 
system, can easily be made to give a closing characteristic approxi- 
mating the shape shown by curve 1, but for reasons which will be 
explained, the hesitation of the discharge must be produced otherwise 
than in the filter. 

144 E. W. KELLOGG [J. S. M. P. E. 


Current for operating a shutter or biasing a galvanometer has in 
all ground-noise reduction systems been derived from a rectifier sup- 
plied with audio-frequency voltage. There are many types and de- 
signs of rectifiers and their performance is by no means equivalent. 
One of the first questions which comes up is the desirability of using 
a full-wave rectifier. The case for half-wave rectification has been 
well put by S. Read, Jr. 17 ** In the first place, it is only the peaks 
in one direction which the shutter has to clear. If the peaks of the 
opposite polarity are smaller, their inclusion in the rectifier will not 
affect the rectified voltage, whereas if they are larger, they will cause 
the shutter to open by an unnecessarily large margin. With pure sine 
waves, or with any truly symmetrical waves, full- wave rectification 
makes it easier to remove the ripple from the rectified current, but 
such ripple as remains is more objectionable in that it is not of funda- 
mental frequency; whereas the ripple which results from half- wave 
rectification is mostly of fundamental frequency, and if the shutter 
executes some movements in response to this ripple current, it simply 
alters the magnitude of the reproduced fundamental tone, but does 
not introduce a tone which was not in the original. 17 Among RCA 
engineers, the arguments for half-wave rectification have, in general, 
prevailed, although it is admitted that there are many situations in 
which full-wave rectification may be desirable. The argument that 
full-wave rectification may give excessive shutter opening would 
not have much weight were speech waves usually symmetrical, or 
random in their asymmetry, but oscillographic studies show that 
speech waves are predominately unsymmetrical and the higher peaks 
are nearly always pressure, rather than rarefaction peaks. 17>18>19 It 
is entirely possible to take advantage of this, and, by properly poling 
the recording system with reference to the microphone, to put the 
shutter on the side that calls for the lesser movements. This reduces 
the magnitude of all the ground-noise reduction problems, including 
ripple, clipping, and shutter noise. 


R. O. Drew, of the RCA laboratory in Camden, some time ago 
made a study of the audibility of various types of opening and closing 
characteristics. He mounted a metal flange around the circumference 
of a low-speed turntable and cut variously shaped curves in the 
flange so that the manner of intercepting a beam of light 1 inch high 


by 0.020 inch wide could be varied by the simple expedient of cutting 
a piece of paper to the desired shape and sticking it to the flange. 
The amplifier gain was set so that full modulation of the light would 
have produced a rather loud signal. A standard RCA reproducing 
channel was employed with a No. 64-A monitoring loud speaker, 
"which has good bass, although perhaps not quite as good as a stand- 
ard theater speaker. This, however, was compensated by our 
proximity to the loud speaker. Results were also checked by listen- 
ing with head-phones of special design which gave good bass. Fig. 
10 shows some of the variations in the masks tried. A number of 
listeners participated. The conclusion was that although there was 
a slight difference in quality of the sound, the general audibility 
depended only upon the maximum slope of the cutting edge. This 
appeared to be substantially independent of the length of the slope. 

/Wclio Inpu-t Vo l-t . 

FIG. 10. (Left} Masks used in testsof audibility. 

FIG. 11. (Right} Desirable type of non-linearity in shutter 


Rounding the corners, as indicated in curves 2 and 4 as compared 
with curves 1 and 3, did not serve appreciably to reduce audibility, 
and if the steepness had to be increased in order to bring the total 
time to the same value, the expedient of rounding the corners was 
definitely harmful. In other words, curve 6 was worse than curve 5. 
It is not often that results can be so simply stated as in this particular 
investigation, and no doubt it will be subject to minor corrections, 
but we believe that it is a reliable guide in design of ground-noise 
systems. We may therefore sum up the requirement of slit unmask- 
ing by saying that the shutter must never move faster than a specified 
velocity, but we do not care how quickly it is accelerated to this 

So long as the maximum shutter velocity is kept below the specified 
value, there would appear to be no reason why the opening should not 
always be at the same rate even though the modulation may not be 
high. It would certainly appear reasonable to permit the highest 

146 E. W. KELLOGG (j. S. M. P. E. 

velocity which will be inaudible until the opening has reached normal 
clearance above the existing modulation. For example, when a 
weak sound starts, the shutter would open as rapidly but not as far 
as it would at the start of a strong sound. The non-linear relation 
between input and output or shutter deflection shown in Fig. 11 is 
a step toward making the opening rate more nearly the same for 
small as for large amplitudes of audio input. 

Whether or not it would be better to employ full- wave rectification 
for steady-state conditions, there is certainly something to be said 

FIG. 12. Examples of speech sounds in which posi- 
tive peaks appear in advance of appreciable negative 

for utilizing waves of either polarity to start the shutter opening. 
Thus, assuming that the shutter and rectifier are so poled that the 
negative peaks are toward the shutter and that these are in general 
smaller than the positive peaks, it is still desirable that the motion 
of the shutter be initiated by the first indication of incoming sound. 
As was brought out in the paper "Starting Characteristics of Speech 
Sounds" 19 there are often positive peaks of appreciable magnitude 
considerably in advance of the negative peaks of comparable magni- 
tude (Fig. 12). We might like to take advantage of this charac- 
teristic and permit the positive peaks to increase the clearance, or, 
in other words, to cause some shutter opening even though the 


ultimate opening may more logically depend upon the height of the 
negative peaks only. 

In discussions of ground-noise systems, we engineers have fallen 
somewhat into the habit of talking about "quick opening" of the 
shutter, when the characteristic we are talking about is not strictly 
a matter of speed but rather of amplitude. For example, to reduce 
clipping during the start of sounds, we may use a circuit or an adjust- 
ment of gain which will cause the shutter to open beyond the bare 
clearance point by a large margin. This, of course, results in faster 
shutter movements because the shutter has farther to go, but funda- 
mentally it is not a part of the shutter timing. Fast shutter timing 
depends upon the filter design, the RC or the LC values, and number 
of filtering elements. These in general are the same for all ampli- 
tudes, and their effects are measured by applying transient impulses. 
On the other hand, the input-output characteristic, which includes 
the effect of the amplifier gain is measured under steady-state condi- 
tions by applying continuous input tones and determining the result- 
ing shutter positions. It is therefore not logical to use such terms as 
"quick opening" or "fast action" to describe an input-output charac- 


In the absence of an anticipating system, we are constantly depend- 
ing upon clearance to prevent clipping. The shutter always requires 
an appreciable amount of time to reach the opening which normally 
corresponds to the amplitude of the incoming waves. Fig. 13 il- 
lustrates the relation between delay, clearance, and clipping. Here 
an audio wave is assumed which increases rapidly in amplitude. If 
the initial clearance is small and if there were no delay in the shutter 
opening, it would follow curve 1 of Fig. 13 (a) and there would be no 
clipping. Delay, however, which inevitably results when we try to 
filter out the audible components from our rectified voltage, causes 
the shutter movement to be represented by curve 2 instead of curve 
1, and thus clipping occurs throughout the period of increasing sound 
amplitude. One way of providing clearance is to permit a substan- 
tial opening even at zero modulation. This will frequently prevent 

The term margin has come to be used to express an excess gain setting (ex- 
pressible in db) above the value which would barely avoid clipping. Clearance 
is used to mean the actual distance by which the shutter image clears the peaks 
of the waves. 



U. S. M. P. E. 

clipping, as illustrated in Fig. 13(6). Such a large residual opening, 
however, tends to defeat the purpose of ground-noise reduction. In- 
creasing the gain of the ground-noise reduction amplifier also reduces 


FIG. 13. Relation between margins and clipping. 
FIG. 14. Fundamental type of ground-noise reduction 


FIG. 15. Exponential opening characteristic and ap- 
proximation to straight-line opening when margin and 

limit are taken into account. 

FIG. 16. Effect of extra filtering upon opening character- 

FIG. 17. Sharp peak type of audio wave and resulting 
rectified and filtered voltages. 

clipping, although as will be seen by comparing Fig. 13(6) with 13(c), 
it does not quite take the place of the larger initial opening so far as the 
first small waves are concerned. Since the gain is a question of ratio 


of input to output, the mere expedient of increasing the gain results 
in excessive clearance for moderately high amplitudes once the full 
opening has been reached (see Fig. 13(c)). Fig. 13(d) shows the clip- 
ping relation with a small residual (or zero modulation) opening, but 
an input-output characteristic of the type illustrated in Fig. 12. This 
is generally regarded as the best practical compromise. The question 
may arise whether a crescendo of the type illustrated in Fig. 13 is 
typical of what is most commonly encountered. This question was 
the subject of an investigation reported by R. O. Drew and the 
writer. 19 We found that in speech sounds there were plenty of 
cases where the amplitude of the first wave was 50 per cent or more 
of that of the ultimate amplitude, but by far the most common type 
was that in which the amplitude increased linearly with time for the 
major part of the growth, with a slowing down of the rate of increase 
as the final amplitude is approached. Such a characteristic of the 
sounds to be recorded justifies writing into our specification for the 
ideal ground-noise system the stipulation that its input-output 
characteristic shall be such as to tend to afford a clearance of a sub- 
stantially constant amount, rather than a margin of constant db 
excess gain. In other words, we should design the system with a 
view to maintaining about the same clearance at all amplitudes (when 
measured by a steady-state test). 

One other characteristic might be asked for, although it may call 
for more complication than its value warrants: namely, that the 
margins (as set by the gain in the amplifier system, for example) shall 
be increased during periods of sound growth, as compared with the 
margin settings during steady-state conditions or during periods of 
diminishing sound amplitude. 

We shall give further consideration to the subject of margins and 
how they can be provided. Up to this point it has been our purpose 
primarily to write specifications for an ideal ground-noise reduction 
system, or to formulate a statement of what we should like it to do. 


Fig. 14 shows, in schematic form, a rectifying and filtering circuit, 
which will serve as a basis of discussion. Conclusions in regard to 
the performance of a circuit such as shown are in general applicable 
to any ground-noise reduction system wherein a rectifier charges a 
condenser and the voltage across this condenser is filtered to produce 
an envelope voltage. For simplicity's sake in many of the arrange- 

150 E. W. KELLOGG tf. S. M. P. E. 

ments to be discussed, we shall omit some details such as providing 
suitable bias for various amplifier tubes. In the design of this circuit, 
the purpose was to provide a current to operate the shutter which is 
proportional to the actual peaks of the rectified waves. The output 
of a power tube (not shown) is stepped down by transformer T to 
provide a low-impedance audio supply for operating the rectifier. 
The loading resistance RI lowers the supply circuit impedance, but 
its most important purpose is to keep the leakage reactance of the 
transformer from interfering with the delivery of very short pulses 
of current. When the audio voltage rises above the value to which 
condenser C\ happens already to be charged, current flows through 
the rectifier D, and charges C\ to substantially the full peak audio 
voltage. RZ and C% constitute a stage of filtering to prevent abrupt 
changes from occurring in the voltage E z which is applied to the grid 
of the output tube, whose anode-cathode current operates the shutter. 
(It has been assumed in our discussion, for the sake of simplicity, that 
the shutter responds instantly to changes in the current which actuates 
it. In actual shutters the inertia of the moving parts is kept as low 
as practicable, so that the characteristic just mentioned is at least 
approximated.) In order that the condensers may discharge and 
permit the shutter to return to the normal when the modulation 
ceases, the discharge resistor Rs is provided. C 2 is made small in 
comparison with C\ so that it will not rob the latter of too much of 
its change. 

Although there are many possible variations in rectifier and filter 
arrangements, the general features of rectifier, condenser to receive 
the charge, filter, and a discharge path, will be found to be common 
to nearly all ground-noise reduction systems, so that what is here 
said of Fig. 14 is, in general, applicable to other systems. 


It is obvious that the voltage E\ produced by the rectifier is not 
suitable to apply directly to the shutter (or rather to the grid of the 
tube which operates the shutter), for it undergoes abrupt changes 
whenever a steep wave-front reaches the rectifier, and would there- 
fore produce noise. The smoothly varying envelope current must be 
derived from the rectified voltage E\ by employing a low-pass filter, 
whose constants are so chosen as to permit rapid enough changes to 
follow the modulation amplitude, but to exclude the objectionable 
small and rapid variations, or rather to reduce these higher frequency 


components to so small a magnitude that they will be practically 
inaudible. In Fig. 14, the filtering is provided by the resistance R z 
and condenser C 2 . 

The most rapid shutter movements would obviously be those which 
result from suddenly applying a full-amplitude audio wave to the 
system. In this case, clipping is inevitable and the shutter should 
move out of the way as rapidly as can be permitted in view of the 
requirements for avoiding audible shutter thump. It will be recalled 
that in our tests for audibility it appeared that motion may begin 
instantly at full velocity and continue until the desired opening is 
reached, as illustrated in curve 1 of Fig. 10. A filter of the type shown 
in Fig. 14 comprising only one resistance and one condenser approxi- 
mates the desired performance. If, for example, an audio wave, such 
as shown in Fig. 15, is applied to the system, the filtered voltage shown 
as E 2 starts instantly to rise at a certain maximum rate which is set 
by the magnitude of the voltage and the values of RZ and Cz. The 
maximum audio voltage should obviously be that which corresponds 
to full modulation of the track. If there is no margin of excess shutter 
opening, the final movements of the shutter will be too slow, as 
shown by the curve EZ, but with a moderate margin the shutter motion 
will be at almost full velocity until the track is wide open as shown by 

In some ground-noise reduction systems a shunt resistance and 
series inductance control the filtered current instead of a series re- 
sistance and shunt capacity. These two types of filter have the same 
fundamental characteristic. 

The filtering obtainable with a single stage employing only one re- 
active element and one resistance may not always prove adequate for 
removing the audible components of the rectified voltage, unless some 
special means, as will presently be discussed, are employed to reduce 
the amount of ripple coming from the rectifier, or, in other words, 
the ripple in jEi. This is particularly likely to be the case if the 
system is designed to make the shutter close rapidly when the modu- 
lation falls. For this reason, filters have been employed which give 
increased filtering, obtained either by using combinations of in- 
ductance and capacity or two or more stages of resistance and capac- 
ity. The effect of either of these expedients is to produce a toe in 
the opening characteristic, as shown in Fig. 16 in which curve 1 
represents the characteristic of a two-stage or double reactance filter, 
as compared with curve 2 for the filter previously described. So far 

152 E. W. KELLOGG [J. s. M. P. E. 

as clipping goes, the delayed starting represented by the horizontal 
offset a is dead loss, and as has already been stated, curve 2 would 
cause no more thump than curve 1. There is nevertheless some excuse 
for employing the extra filtering elements. The problem of avoiding 
objectionable noise is not solely one of limiting the maximum rate of 
opening, although this is the only part of the problem where there 
seems to be a necessary compromise. Under steady-state conditions, 
the rectified voltage EI, of course, contains components of many fre- 
quencies. When the characteristics of the ear and those of the re- 
producing system are combined, the result is that loudness increases 
with frequency at a very rapid rate in the low-frequency range. The 
attenuation of the higher frequencies by a single stage of resistance- 
capacity filtering does not increase fast enough to compensate for the 
increased audibility of the high-frequency components. On the other 
hand, other types of filters can be made to discriminate more rapidly 
against the high frequencies. The easiest way to get complete sup- 
pression of audible components is to use the extra filtering. Were 
there no other way of avoiding the audible noise, we should have to 
resign ourselves to the use of the extra filtering with its consequent 
delay in shutter opening. It is much better, however, to reduce the 
disturbance at its source, and there are a number of factors by which 
the ripple voltage, which has to be filtered out, can be reduced in 
magnitude. When this is done, it is quite possible to meet the noise 
requirements with the single-stage of filtering. 


The voltage (or current) at the output end of a low-pass filter de- 
pends upon the average value at the input end rather than upon its 
peak value. Thus in Fig. 17, EI and E 2 have the same average value. 
Therefore, it is the average value of E\ (rather than the maximum 
value) which we wish to make as nearly as possible equal to the peak 
of the audio wave applied to the rectifier. In order to meet this 
requirement, the input circuit and rectifier must be of low enough 
resistance so that the condenser will reach full charge during the time 
which the audio wave applies this voltage. This is not very impor- 
tant in the case of sine waves, for if they are of low frequency the 
voltage is sustained for a considerable period of time; while if they 

*The term peak-reading is borrowed from voltmeter specifications to describe 
a system in which the shutter opening depends upon the magnitude of the peaks 
of the audio waves, rather than on their average or their rms value. 


are of high frequency and the condenser does not receive full charge 
during the first cycle, the charging will be completed by more waves 
following in quick succession. In actual recordings, however, waves 
of the type shown in Fig. 17 are extremely common, not necessarily 
in the exact form indicated, but waves whose peaks are of very short 
duration and in which an equally high peak does not occur until a 
whole cycle (of fundamental frequency) later. Waves of this type 
present the greatest difficulty in providing peak-reading character- 
istics. The first requirement for a peak-reading system is a low-re- 
sistance supply circuit and rectifier, in order that EI may reach the 
full peak value in a single half -wave of a high-frequency voltage. The 
input circuit impedance must also be free from appreciable inductive 
reactance. The next requirement is that there shall not be too much 
discharge of condenser C\ between the moments of charging. Ob- 
viously, this requirement is met when the discharge or closing rate 
is made very slow, for example, by using a high value of RS in Fig. 14. 
With variable-area recording a slower discharge has been acceptable 
than in the case of variable-density recording. Values of C\ and R$ 
have been widely used for area recording, which give a time constant 
of 0.045. Even with this timing, however, C\ could lose about 33 
per cent of its charge between peaks spaced 10 milliseconds apart 
(corresponding to a 100-cycle fundamental). With faster closing, as 
called for in density recording, peak-reading properties are further 
impaired. This criticism does not apply particularly to the circuit 
of Fig. 14. The only type of circuit which will give true peak readings 
is one in which the rectifier is virtually unloaded and the condenser 
does not discharge appreciably during one cycle of fundamental 
frequency. Therefore in the design of a ground-noise reduction sys- 
tem, it should be the aim to come as nearly to meeting these two con- 
ditions as is practically possible. A heavily loaded rectifier will give 
something between an average and an rms value of the applied volt- 
age, and thus far from the peak reading. 


As has been stated, ground-noise reduction systems are invariably 
operated with a certain margin of excess gain, usually adjusted by a 
potentiometer volume control in the audio amplifier which feeds the 
rectifier. It is customary to set the margin by applying a 1000 -cycle 
wave of specified amplitude and adjust the ground-noise reduction 
gain until the shutter is opened to a specified point. A margin of 6 

154 E. W. KELLOGG [j. s. M p. E. 

db would mean that any change in the amplitude of the audio waves 
as recorded on the film would result in a shutter image movement of 
exactly twice the amount. If, for reasons already discussed, the 
shutter movement does not bear a linear relation to the audio ampli- 
tude, then we can not speak of so many db margin except in connec- 
tion with a specified test- tone level. The margin serves a twofold 
purpose. It prevents clipping in case of an increase in audio ampli- 
tude, provided the increase is less than the clearance which this 
margin has produced; and in the second place, margin is required 
under steady-state conditions because ground-noise reduction systems 
are in practice far from truly peak-reading, adjustment for a large 
margin on sine waves, for example, being only sufficient to give bare 
clearance for a wave of equal peak height but having the shape shown 
in Fig. 17. Peak-reading properties are just as important for the 
purpose of maintaining the margins during crescendos as they are for 
maintaining them during a steady state. Consider, for example, a 
series of waves of increasing amplitude but of comparatively low 
fundamental frequency. If there is failure to charge the condenser 
completely during a peak or if it discharges between peaks, the in- 
tended margin to allow for crescendo is lost, and again the only cure 
is to provide still more margin of gain. We thus see that poor peak- 
reading properties must be compensated by substantial increases in 
margin adjustments. The margin may not be excessive with certain 
types of wave, while with others it will be decidedly more than neces- 
sary. The obvious penalty for employing more margin than neces- 
sary is an increase in ground-noise. Since ground-noise is scarcely 
a problem when the modulation is high, engineers are inclined to 
dismiss the erratic and often excessive margins as of no consequence, 
to set the margins high enough to prevent excessive clipping, and 
consider the system to be working satisfactorily. Such a viewpoint, 
however, ignores the most important factors. It is true that some 
excess of clear film when the modulation is high does little harm, but 
what about when the modulation stops? The shutter has farther to 
go, and a faster rate of closing is necessary to prevent audible ground- 
noise after the modulation. This aggravates ripple, making more 
filtering necessary, and also makes the peak-reading properties worse. 
Again during opening, if the shutter moves farther than necessary, 
it must also move faster than necessary. This results in more trouble 
from shutter thump and more danger of noise due to small move- 
ments of the shutter. The cure for the last trouble is again more 


filtering, with consequent increased delay of shutter action which 
must be compensated by larger margins in order to prevent excessive 
initial clippings. Poor peak-reading properties (by which is here 
meant inconsistent relationship between peak height and shutter 
opening) thus initiates a vicious circle of corrections for faults, with 
an exaggeration of all the troubles to which ground-noise reduction 
systems are heirs. 

We shall next consider some specific circuits and proposals for 
improving the action of ground-noise systems. 


An ingenious and interesting proposal for improving filtering, while 
still permitting as quick operation as may be .desired, consists in 
modulating a high-frequency current by means of the audio waves. 
This is done in a balanced modulator which eliminates the carrier 
frequency. A sharp filter then removes one of the side-bands. The 
remaining side-band is a measure of the magnitude of the audio 
frequency, and, if rectified, will give a steady current proportional to 
the audio amplitude. Thus if we start with a 100-cycle audio input 
and use this to modulate a 20,000-cycle carrier, there would result 
voltages of 19,900 and 20,100 cycles as side-bands. The magnitude 
of the side-bands would be directly proportional to that of the original 
100-cycle current. Only one of the side-bands, the 20, 100-cycle one, 
for example, is applied to the rectifier. A low-pass filter having a cut- 
off frequency of several thousand cycles would serve to remove the 
high-frequency component after rectification, and such a filter would 
not need to produce appreciable delay. 

On the basis of a pure sine wave input, the system just described 
appears to be perfect. As soon, however, as we assume that the audio 
waves contain several component tones, much of the advantage of 
the single side-band system disappears. Let us assume, for example, 
an audio wave having equal components of 100 and 200 cycles. The 
upper side-bands would then be of 20,100 and 20,200-cycle frequency 
and would be equal in magnitude. The sum of these two voltages 
would be a high-frequency voltage modulated 100 per cent at a 100- 
cycle modulation note. Rectification would result in a continuous 
current proportional to one of the high-frequency components, 
superimposed on a 100-cycle wave having a maximum value the 
same as that of the continuous current, thus modulating it by 100 
per cent. It would therefore be necessary to filter out the 100-cycle 

156 E. W. KELLOGG ' [J. S. M. P. E. 

component by means of a low-pass filter comparable with that re- 
quired for one of the direct rectifier systems already discussed. It 
appears that in this case nothing at all has been gained by the com- 
plication of modulating the high frequency. With many audio wave- 
forms, the result of rectifying the upper side-bands would probably be 
intermediate between the results with the pure sine wave and the 
rather bad case just discussed. So far as the writer has been able to 
judge, it is doubtful whether with complex waves the single side-band 
system would offer enough benefit to justify the circuit complexities. 
The system just described must not be confused with the systems 
employing carrier current subsequent to the production of a filtered 
envelope voltage. 20 Such systems have been in successful use, but 
in these the carrier-purrent feature may, so far as the discussions in 
this paper are concerned, be regarded simply as serving as an ampli- 
fier to provide an output current proportional to the filtered voltage. 


Referring again to Fig. 17, it is evident that the amount of ripple 
in the voltage EI which has to be filtered out, and also the drop below 
peak value, are reduced by slow discharge. In attempting to formu- 
late how we should like a ground-noise reduction system to perform, 
we specified that we should like the closing to be represented by curve 
1 of Fig. 9 rather than curve 2. If C\ of Fig. 14 discharges in the 
normal exponential manner as illustrated by curve 2 of Fig. 9, the 
filter itself will cause E 2 to discharge according to a curve something 
like no. 1 of the same figure, but this is not what we want or intended 
when formulating the specification. Once a fraction of the charge 
has been lost from Ci the circuit will not continue to hold the shutter 
open for the arrival of the next peak. No appreciable discharge of 
either condenser should take place until the lapse of a certain time 
after C\ received the charge. The voltage across C\ should thus be 
represented by curve 1 of Fig. 9. It is not necessary that the entire 
discharge shall be at a low rate (such as would be caused by using a 
high resistance for R z of Fig. 14), in order to give the peak-reading 
and low-ripple properties to the ground-noise system. If a full period 
of the audio wave passes and a new peak does not arrive, this is good 
enough evidence that discharge at normal rate should begin so that 
the envelope curve will follow the modulation down. 

The advantages of delayed discharge in reducing ripple apply in an 
even more important way to the opening characteristic than to steady- 


state. Fig. 18 shows a crescendo of a wave of low fundamental fre- 
quency, and the rectified but unfiltered voltage EI derived from it, 
first, with a moderately high initial discharge rate (curve 1) such as 
would be given by a simple resistance discharge path as in Fig. 14 
and fairly rapid closing; and, secondly (curve 2), with slower closing 
or delayed discharge, such that C\ would not discharge appreciably 
between the charging peaks. In either case it will be noted that EI 
rises by steps instead of uniformly as would be wanted for minimum 
noise. In other words, superimposed upon the desired uniformly ris- 
ing voltage is a strong ripple, much larger than the ripple which will 
be produced after full amplitude is reached. This ripple is much 
worse in curve 1 than in curve 2. When the objective is to suppress 
noise to the point where it is inaudible a few db difference in level may 
become very important. During crescendos not only is the ripple 

FIG. 18. Step effect on charging and exaggeration of 
same by discharge between input peaks. 

larger than during a steady state, but the fact that it is superimposed 
upon a continuous change which is itself as rapid as can be tolerated, 
makes the ripple at this time more objectionable than an equal ripple 
under steady-state conditions. The ripple in the filtered voltage 2 
will, of course, be much less than that in EI as shown in Fig. 18, but 
whatever increases the ripple in EI will increase it in 2 . Two factors 
have undoubtedly contributed to toleration of unnecessarily large 
crescendo ripple: (1) the quality of the sound resulting from a short 
train of low-frequency waves is so similar to that produced by the 
necessary shutter change that the latter may often be blamed for 
what is in part due to the former, and (2) crescendos are so often ac- 
companied by clipping that either the clipping distortion masks the 
low-frequency disturbance, or else all the faults are lumped together 
and treated as inevitable. 

158 E. W. KELLOGG \j. s. M. p. E. 


A number of arrangements have been studied with a view to provid- 
ing a discharge characteristic approximating that of curve 1, Fig. 9. 
It is the writer's belief that further improvements in ground-noise 
reduction systems will involve employing some such expedient, the 
problem being to select the most promising of the suggestions con- 
sidered, and then to prove on actual test that the improved perform- 
ance is worth the added complication. In judging on the basis of 
test whether a complication is justified, the writer would like to 
emphasize a point of view which he believes to have been frequently 
vindicated. If a device theoretically performs better than another 
(without sacrifice of such factors as reliability), and if the theory is 
recognized as sound, it may often occur that tests on a limited scale 
will fail to show an impressive advantage, but the better functioning 
device should be employed on the simple probability that ultimately 
its superiority will show up in the quality of the product. Ground- 
noise reduction itself is an example of such an experience. Robinson's 
first tests convinced him and some of his associates that the principle 
was worth while, but others were not convinced. Ultimately its 
importance became established. 

Fig. 19 shows a circuit proposed for the purpose, not of preventing 
the discharge of condenser Ci, but of renewing the charge at short 
intervals covering a total period which might be of the order of ten 
to twenty milliseconds. The transmission line shown would, of 
course, be damped at the end to prevent reflections, and would have 
to be of low enough impedance not to be appreciably loaded by the 
rectifiers. There are, of course, numerous modifications involving 
the same principle. 

An inductance in series with R$ of Fig. 14 would delay discharge, 
producing a curve with a shoulder. On first thought this appears to 
offer a possible solution for the problem of delaying discharge. Un- 
fortunately, however, inductance acts in the manner described only 
when there was no previous current flowing through it. Under steady- 
state conditions, it is not of any appreciable help toward giving peak- 
readings. The inductance would, however, serve a useful purpose 
in that it would reduce the amount of discharge during crescendos, 
thereby increasing the margin. 

Fig. 20 shows a modification of the circuit of Fig. 14 in which an 
additional condenser C 3 is charged simultaneously with C\. Assuming 
that both these condensers become charged to the full peak voltage, 

Feb., 1941] 



Ci does not begin to discharge rapidly until the voltage across C$ 
has fallen considerably. This circuit produces a result in the right 
direction, but if the resistances are so chosen that d will be discharged 
down to about 10 per cent of its original charge in the same time as 


FIG. 19. Circuit proposed by the writer for providing successive 


FIG. 20. Two-stage discharge circuit. 
FIG. 21. Circuit of S. Read, Jr., employing grid-controlled 

FIG. 22. Differential circuit of H. I. Reiskind. 

in the circuit of Fig. 14, the shoulder which the Fig. 20 circuit pro- 
duces in the discharge of C\ does not extend as far as we should like. 
Fig. 21 shows an arrangement proposed by S. Read, Jr.,** employ- 
ing an expedient first suggested by G. L. Dimmick:** namely, the 
deriving of a control voltage from the secondary of a transformer 
whose primary is in series with the output tube. Dimmick showed 

**RCA Manufacturing Co,, Camden, N. J. 

160 E. W. KELLOGG [j. S M. P. E. 

that the voltage induced in the transformer secondary could be made 
effectively to oppose the discharge of capacitor C\ by connecting the 
transformer winding directly in series with R s . This opposing volt- 
age (being proportional to the rate of change of current through the 
output tube) is directly proportional to the drop across R 2 and per- 
sists, but with diminishing value, until Cz has become charged to the 
same voltage as C\. This helps materially when the audio voltage 
is first applied or during crescendos, but under steady-state condi- 
tions the retarding voltage becomes too small to do much good in 
the way of reducing ripple. Dimmick worked out several arrange- 
ments for increasing the opposing voltage effect without permitting 
it ever to raise the condenser voltage EI above the audio peak value. 
Read employed a triode through which the discharge takes place, 
and used the transformer voltage to bias the tube to cut-off. By 
using a high-"mu" tube and a large number of turns on the trans- 
former, the discharge can be blocked so long as there is even a small 
rate of increase in the voltage across Cz. The biasing voltages will be 
properly timed, for the moments of maximum rate of rise of voltage 
across Cz occur just after C\ has received a charge. 

It has already been stated that a second stage of resistance-ca- 
pacity filtering gives a shoulder on the discharge curve such as curve 
1 of Fig. 9. Mathematical analysis of such a circuit shows that the 
equation for the discharge curve takes the form of the difference be- 
tween two exponential decay curves, one having a large initial value 
and a slow time-constant and the other of smaller initial value and 
more rapid discharge. The initial slopes of the two curves are of the 
same value, but being of opposite sign, the resultant or difference 
curve starts horizontally, but the effect of the smaller component 
soon dies out, and the remainder of the discharge practically coin- 
cides with that of the larger circuit. Mr. H. I. Reiskind** has experi- 
mented with a circuit shown in Fig. 22 in which the total voltage E\ 
is equal to the difference between the voltages across two condensers, 
both charged simultaneously and oppositely but to different voltages, 
the condenser charged to the lower, subtractive voltage being shunted 
by a resistor which discharged it more rapidly than the other. With 
this arrangement, greater freedom in choosing constants was ob- 
tained and the desired discharge curve could be produced without 
some of the limitations imposed by other methods. 

Fig. 23 is from British patent No. 505,011 to Karl Schlegel. The 
letters designating such condensers as have counterparts in Fig. 14 






i r -U JLL 

^ Q 

.. B 

! i, 



* k* 2 



5 T 


f. 3 

~c ~ 






I t:+ ^i 

Audio Input 


FIG. 23. Circuit of Karl Schlegel, Brit. Pat. No. 505,011. 
FIG. 24. Circuit of J. B. Gehman, using gas-discharge triodes. 
FIG. 25. Rectox recording unit of G. L. Dimmick and type of 
response curve produced by it. 

162 E. W. KELLOGG [j. S. M. P. E. 

have been changed from those shown in the patent to correspond more 
closely to the designations already used in this paper. Thus C\ is 
charged to the peak value of the audio wave, and the voltage across 
it is designated as E\. After smoothing by the filter Rz and Cz, the 
voltage (now designated as Ez) is impressed upon the grid of the out- 
put tube N. The discharge path in this circuit is across Cz. There 
is, however, no discharge path of fixed resistance. The rectifier 
element D* is so poled that it simply sets an upper limit to the voltage 
EZ. The only discharge path is through rectifier D s to conductor B 
which is charged by the extra rectifier D\ to a voltage 3, higher than 
E\. Rectifier element D s is so poled that it prevents Cz from receiving 
any charge from the high-voltage rectifier DI, but does permit Cz 
to begin to discharge after 3 has fallen to a lower value than EZ. 
(Di is another voltage limiter.) Instead of a simple resistance for 
discharging condenser C 3 (which is charged to the voltage 3 ) the 
pentode U is employed. It is well known that a pentode has a con- 
stant-current characteristic, and the purpose is evidently to cause the 
discharge curve to take the form of a straight line, the voltage falling 
at a constant rate, instead of along a sagging curve of the exponential 
type. The patent does not show any provision for making this con- 
stant-current effect work down to complete discharge or zero plate 

Fig. 24 shows a circuit developed by J. B. Gehman.** The shutter 
current is supplied by several output tubes T in multiple. Only two 
are shown in the figure, for illustration, but about five circuits such 
as shown would be employed. Each output tube is adjusted to carry 
a desired fraction of the total current. The tubes Gi, Gz, etc., are of 
the gas triode type. The audio input voltage is applied across the 
potentiometer P, and the grid of each gas tube is connected to a 
suitable point on the potentiometer, such that they are rendered 
conducting at successively higher voltages across potentiometer P. 
When the grid of any tube, GI, for example, reaches the tripping 
potential, the tube at once carries a large current, completely dis- 
charging condenser C\. Since the plate voltage drops to zero, the 
conduction ceases as soon as C\ is discharged. Thereupon, C\ im- 
mediately begins to charge again through resistance RI. The output 
tube 7i is so biased that its plate current is completely cut off when 
the voltage across Ci is below a specified value. Thus the plate cur- 
rent of tube 7\ drops from the normal level which it has during periods 

**RCA Manufacturing Co., Camden, N. J. 


of no modulation, to zero, remains at zero until C\ has charged to the 
specified value, and thereafter rises and approaches normal zero 
modulation value again at a rate determined by C\ and RI and the 
supplied voltage. Filtering is provided by shunting the shutter 5 
by a capacitor C f . 

At zero modulation, the shutter current is the sum of the plate cur- 
rents of all the tubes. A low-level audio wave will trip one gas tube 
and decrease the current of the corresponding output tube to zero. 
The next higher level (chosen by the position of the next contact on 
the potentiometer P) will trip the first and second tubes. Still higher 
level will trip three tubes, and so on. The filtering must be so ad- 
justed that excessive thump will not occur if all tubes are tripped at 
once. If they are tripped successively by the audio wave, the fall in 
total current will be less rapid. Whatever value of current is estab- 
lished by the level of modulation and the number of tubes biased to 
cut-off, this value will be retained until a predetermined interval has 
elapsed after the passage of the last audio wave of sufficient magnitude 
to trip these tubes. If, after a certain group of tubes has been tripped, 
another audio wave of the same amplitude occurs, the same gas tubes 
will again conduct current and discharge their condensers. Recovery 
must then start over again. This is true, regardless of whether the 
previous recovery had progressed far enough to cause any of the 
output tubes to carry plate current or not. If the new audio waves 
reach the input circuit before conduction through the output tubes 
begins, then there is no ripple whatever in the output or shutter 
current, for each output tube in question continues completely cut- 
off. Ripple begins to appear when the interval between audio waves 
is greater than the period of complete cut-off, which interval is fixed 
by the design or adjustment of the system. Ripple becomes pro- 
gressively greater as the frequency is lowered below this point, but 
would in no case be worse than that of a ground-noise system of the 
usual type having comparable total discharge time. The shutter 
opening, instead of being a continuous function of input level, has a 
fixed set of values. About five output tubes would make the steps 
small enough to maintain the margins within suitable limits. 

The circuits just described illustrate a few of the expedients being 
studied. In most of the illustrations the circuits have been simplified 
down to the bare essentials for explaining the general principles, the 
actual practical circuits having numerous elements which do not 
appear in these sketches. 

164 E. W. KELLOGG (j. s. M. P. E. 


The provision of nearly constant clearance requires that the gain 
or ratio of output to input shall be large for low-level inputs, and 
decrease as the level is raised. This calls for decidedly non-linear ele- 
ments, operating either on the audio waves, or on the rectified voltage 
or current, or in the design of the shutter itself. In the arrangement 
illustrated in Fig. 14, modulation reduces the plate current from an 
adjusted initial value to zero current at full modulation. The toe of 
the plate current vs. grid voltage curve of the output tube tends to 
give the system an arched curve of shutter displacement vs. input 
voltage, somewhat resembling Fig. 12. B. Kreuzer** proposed to 
produce a more pronounced effect of this kind by employing expo- 
nential output tubes. Since it was felt that a still steeper initial rate 
followed by a sharper break was desirable, G. L. Dimmick proposed 
controlling the shape of the curve by means of a biased loading circuit 
with copper-oxide rectifiers as illustrated in Fig. 25. This gives an 
input-output characteristic which is very steep at the start, followed 
by a much slower subsequent increase. Thus small modulation is 
sufficient to produce a substantial margin of shutter clearance, and 
thereafter the clearance may remain about constant, or increase 
slightly until full opening is approached. Tests have shown this to 
be definitely helpful in reducing initial clipping. 


It has been proposed to delay the recording of the sound for ten 
to twenty milliseconds, or perhaps more, after it has been picked up 
by the microphone. On the other hand, the audio voltage would be 
impressed on the ground-noise reduction rectifier without delay. This 
would give the shutter a corresponding time to move out of the way 
in anticipation of an increase in recorded amplitude. One of the 
obvious expedients for providing such delay would be to use a speaker 
unit and a microphone connected through a tube, perhaps twenty 
feet long, and record the output of the microphone instead of the 
original sound. Another delay system consists in recording on mag- 
netic tape or wire, and reproducing by means of a pick-up a suitable 
distance from the recording magnet. A method of delaying sound for 
a short interval by illuminating a moving luminescent surface with 
modulated light, and picking up photoelectrically a short distance 
away, has been described by Goldmark 21 who has used this principle 


to produce reverberation effects. Another method would employ 
bound electrostatic charges on a moving surface. 22 Recording on 
any known medium is of course possible, but the types of record 
which can be erased and leave the carrier ready for fresh recording 
are obviously preferable. 

It will be noted that the audio waves to be delayed are not the 
ones which control the ground-noise reduction system, for which 
purpose considerable distortion might be tolerated, but are the 
waves which must be impressed on the recording device. Therefore, 
distortion is not to be tolerated. It is the unwillingness on the part 
of engineers to introduce an element which could even threaten to 
produce distortion which has no doubt been largely responsible for 
failure to employ one of the expedients just described. Since it is the 
devices which convert waves to mechanical or other forms and convert 
them back to electrical waves, which are responsible for most distor- 
tion, and the limitations of most recording mediums (wax, lacquer, 
magnetic materials, etc.) result in noise, the ideal method of providing 
delay would be one which involved no conversions, as, for example, an 
electrical transmission line. It should be possible and not difficult 
to build such a line which would produce the necessary delay without 
causing measurable distortion. The requirements, however, are such 
that it would be an expensive piece of equipment. Let us assume, 
for example, that the audio channel must be good to 10,000 cycles. 
The line would almost of necessity be made with lumped capacities 
and inductances and would thus act as a low-pass filter. The cut-ofT 
frequency for a low-pass filter is reached when there are TT coils per 
wavelength, but serious phase-distortion occurs before the cut-off 
frequency is reached. It would therefore not be safe to count on less 
than four or five coils per wavelength, and there is question about 
this figure. If the sound is to be delayed 0.02 second, the length of 
the line would have to be 200 wavelengths of 10,000-cycle current; 
therefore, there would have to be at least. 1000 coils and condensers. 
The attenuation will be very high unless the elements are of very low 
loss, and the series and shunt losses must be closely balanced in 
order not to cause decided distortion. The electric transmission line 
as a means of producing delay may find a place, but it is not surprising 
that it has not come into use. The acoustic delay tube seems to be 
about the nearest approach to a satisfactory device, but the writer is 
inclined to think that sound departments are right in not taking any 
chances of distortion or of the failure of any device to function per-. 

166 E. W. KELLOGG [j. s. M. P. E. 

fectly, tolerating rather the amount of clipping which the ground- 
noise system, in the absence of anticipation, causes. 

In original recordings, the direct positive system 13 permits antici- 
pation. The blackening for ground-noise reduction purposes is ac- 
complished by an auxiliary light-beam which operates on the film 
after it has passed the recording light-beam. This system has much 
to commend it as a means of making original recordings. So far as 
the writer can see, it can not be applied to variable-density recording, 
except in conjunction with electric compressors, which would raise 
the modulation when the fogging beam is employed. Original re- 
cordings can also be made without any clipping troubles by adopt- 
ing class 23 or class A-B 24 push-pull recording. The class B system 
was adopted several years ago for single-film newsreel work, 25 one 
of the important considerations being that low ground-noise is ob- 
tained without the necessity of any ground-noise reduction equip- 
ment, and more recently has been used for high-quality studio 
recording. 26 

When we come to re-recording, it would appear that anticipation 
can be very easily obtained by the simple expedient of employing a 
second reproducing system with suitable longitudinal offset. The 
first reproducing system operates the ground-noise reduction and the 
second reproducing system provides currents for recording. There is 
little question that this system would have been adopted long before 
this were the re-recording operation a simple matter of a single re- 
producer. The elaborate systems of re-recording employed in the 
motion picture studios, involving up to a dozen film-phonographs, 
each contributing a selected portion of the program, so complicates 
the problem of providing anticipation that it has, in most cases, been 
regarded as impracticable. Each of the contributing film-phono- 
graphs would need to have a second reproducing system. The entire 
mixing system would have to be in duplicate, with single control, 
and a duplicate amplifier and volume-control system would be needed 
for operating the ground-noise reduction system of the recorder. 
Even this complication might be accepted if the benefits of antici- 
pation were fully appreciated. This, however, comes down to a 
matter of opinion. We hear many faults in reproduction and can 
not always ascribe them to specific apparatus. The elimination of 
all disturbances due to quick opening and all distortion due to initial 
clipping would seem to the writer to be worth a considerable price 
in added equipment, particularly if .the original recordings are made 


in such a manner that undipped sound-tracks are available as originals. 
It appears to the writer that the above-described re-recording system 
with complete duplication of reproducing channels might be avoided 
while still obtaining a large part of the benefit of anticipation. In 
motion picture work, it is common for a single film-phonograph to 
supply practically all the dialog, while the others are supplying music 
and other effects which are quite secondary both in importance and 
amount. The dialog might all be carried by a single film-phonograph 
with double reproducing optics. The output from its auxiliary optical 
system would be supplied to an extra ground-noise reduction amplifier 
at the same gain as the output from its regular optical system. The 
outputs of the regular and the auxiliary ground-noise reduction ampli- 
fiers would be combined through rectifiers in such a way that the 
shutter opening corresponded to whichever of the two systems pro- 
duced the higher voltage at the moment. This would not increase any 
of the steady-state margins, nor would it in any way jeopardize the 
recording, since the regular ground-noise reduction would function 
exactly as at present. In fact, about the only time that the auxiliary 
system would affect the shutter movement would be when the shutter 
ought to be opening but the regular signal has not yet reached it. 
The advantages of the anticipation system are : 

(2) Avoidance of all initial clipping. 

(2) Permitting less rapid opening and thereby avoiding dangers of audible 

(5) Ability to reduce margins (because present margins must be exaggerated 
in order to allow for a surprise factor). 

(4) Permit quicker closing. Present closing speeds are dictated by the neces- 
sity, on the one hand, of suppressing ground-noise as promptly as possible after 
modulation has fallen; and on the other hand, the danger that every time the 
track closes down there is likelihood of clipping on the next rise of modulation. 
Anticipation by eliminating the latter danger, makes it practicable to follow the 
modulation down as rapidly as desired. 


1 ROBINSON, L. T.: U. S. Pat. No. 1,854,150. 
8 HEWLETT, C. W.: U. S. Pat. No. 1,835,872. 

3 HANNA, C. R.: U. S. Pat. No. 1,888,724. 

4 MCDOWELL, H.: U. S. Pat. No. 1,855,197. 

5 TOWNSEND, R. H., CLARK. L. E., AND MCDOWELL, H.: "Ground-Noise 
Reduction RCA Photophone System," Tech. Bull. Acad. Mot. Pict. Arts & Sci. 
(Feb., 1931). 

6 SILENT, H. C., AND FRAYNE, J. S. : "Western Electric Noiseless Recording," 
Trans. Soc. Mot. Pict. Eng., XVIH (May, 1932), p. 551. 

168 E. W. KELLOGG [j. s. M. P. E. 

7 BATSEL, C. N., AND KELLOGG, E. W. : "A Shutter for Use in Reduction of 
Ground-Noise," /. Soc. Mot. Pict. Eng., XVII (Aug. 1931), p. 203. 

8 KREUZER, B.: "Noise-Reduction with Variable-Area Recording,"/. Soc. 
Mot. Pict. Eng., XVI (June, 1931), p. 671. 

9 DIMMICK, G. L., AND BELAR, H.: "Extension of the Frequency Range of 
Film Recording and Reproducing," /. Soc. Mot. Pict. Eng., XIX (Nov., 1931), p. 

10 DIMMICK, G. L. : "Galvanometers for Variable- Area Recording/' /. Soc. 
Mot. Pict. Eng., XV (Oct., 1930), p. 428. 

11 DIMMICK, G. L.: "The RCA Recording System and Its Adaptation to 
Various Types of Sound-Track, " /. Soc. Mot. Pict. Eng. t XXIX (Sept., 1937), p. 

12 SACHTLEBEN, L. T.: "Characteristics of Photophone Light Modulation 
System," /. Soc. Mot. Pict. Eng., XXV (Aug., 1935), p. 175. 

13 DIMMICK, G. L., AND BLANEY, A. C. : "A Direct Positive System of Sound 
Recording," /. Soc. Mot. Pict. Eng., XXXIII (Nov., 1939), p. 479. 

14 ROBINSON, L. T. : U. S. Pat. No. 1,935,417. 

15 GREENTREE, C. D.: U. S. Pat., No. 1,971,181; KELLOGG, E. W.: U. S. Pat. 
No. 2,096,811. 

16 STEVENS, S., AND DAVIS, H.: "Hearing," John Wiley & Sons, New York 

17 READ, S., JR.: "A Neon Type Volume Indicator," J. Soc. Mot. Pict. Eng., 
XXVIII (June, 1937), p. 633. 

18 HATHAWAY, J. L. : "Microphone Polarity and Overmodulation," Electronics 
(Oct., 1939), p. 28. 

19 DREW, R. O., AND KELLOGG, E. W.: "Starting Characteristics of Speech 
Sounds," J. Soc. Mot. Pict. Eng., XXXTV (Jan., 1940), p. 43; /. Acoust. Soc. of 
Amer. (July, 1940), p. 95. 

Weight Sound Recording System," /. Soc. Mot. Pict. Eng., XXXIII (Oct., 1939), 
p. 449. 

21 GOLDMARK, P. C., AND HENDRiCKS, P. S. i "Synthetic Reverberation," J. 
Soc. Mot. Pict. Eng., XXXIII (Dec., 1939), p. 635. 

22 KELLOGG, E. W.: U. S. Pat. No. 2,197,050. 

23 DIMMICK, G. L., AND BELAR, H.: "An Improved System for Noiseless 
Recording," /. Soc. Mot. Pict. Eng., XXIII (July, 1934), p. 48; DIMMICK, G. L.: 
"Recording of Eternal Road," Electronics (April, 1937). 

24 CARTWRIGHT, C. H., AND THOMPSON, W. S.: "Class A-B Recording Optical 
Systems," /. Soc. Mot. Pict. Eng., XXXIII (Sept., 1939), p. 289. 

25 DIMMICK, G. L., AND SACHTLEBEN, L. T.: "An Ultraviolet Push-Pull Re- 
cording Optical System for Newsreel Cameras," /. Soc. Mot. Pict. Eng., XXXI 
(July, 1938), p. 87. 

26 BLOOMBERG, D. G., AND LOOTENS, C. L. "Class B Push-Pull Recordings for 
Original Negatives," /. Soc. Mot. Pict. Eng., XXXIII (Dec., 1939), p. 664. 


DR. FRAYNE: Which of the following three elements are the limiting factors 
in obtaining the maximum amount of noise-reduction: (I) the noise-reduction 


circuit; (2) the type of modulator; or (5) the film medium, in their order of 
importance? Where do you stop 10 db, 20 db? 

MR. KELLOGG: In the variable-area system the limit is set by how narrow a 
clear area we can employ when the modulation is low. In our Class B system, 
we can reduce the residual opening to such a point that there would be no appre- 
ciable advantage in going any further. Practically the entire film is black. In 
a standard variable-area recording, we have over a long period of time worked 
with about a 2-mil width of clear film. 

I do not know whether I had better try to answer the question as to what 
would make the most difficulty in going further. May I reciprocate by asking 
how far you can go on the density system? 

DR. FRAYNE: In push-pull systems it has been found feasible to use up to 
20 db of noise-reduction. Common practice with single-track recording is to 
use not more than 10 db. 

MR. SCOVILLE: In the circuits shown by Mr. Kellogg the filters were com- 
posed mainly of resistance and capacity elements. In Western Electric variable- 
density systems we found that type of filter inadequate. There may be a good 
reason for this. With variable-area recording the release time is usually quite 
long, whereas with the variable-density it is relatively short. In consequence 
of that we found it necessary for variable-density recording to have the maximum 
filtering for a given set of timing conditions. That has always called for very 
carefully designed inductance and capacity filters with rather critical adjustment 
of operating time. 

I believe Mr. Kellogg stated that with filters using inductance and capacity 
he found an initial delay in the attack period that was undesirable. We find it 
possible to get around that by the use of a modified Af-derived type of filter, 
which in effect tends to short-circuit the inductance in the initial part of the 
attack period and thus permits a sharp build-up of bias current. 

MR. KELLOGG: Have you found that the JW-derived type of filter gives you 
an almost instantaneous start? Do you know of any experience that either 
confirms or contradicts what I have reported from our tests: namely, that the 
velocity of movement is the determining factor in audibility, and that the sharp 
corner did not seem to do any harm? We did not feel the test was necessarily 

MR. SCOVILLE: The velocity and amplitude of movement are probably the 
chief factors. It is surprising to make comparative tests of filtering for several 
different types of filters, all, however, having the same attack and release times, 
and find how widely the effectiveness of filtering varies. The least efficient type 
of filter in our experience has been the simple capacity-resistance type. 

MR. KELLOGG: My advocacy of the simple capacity-resistance filter is based 
upon the assumption that we would be able to reduce the necessity of filtering 
by some of the measures I mentioned in the latter part of the paper. It seems 
to me that is the direction in which we should like to go. I check your point 
that it is the relatively slow closing in the variable-area system that enables us 
to get away with the simpler filter. 

MR. LIVADARY: Mr. Kellogg stated that he had investigated thoroughly the 
initial rate of increase of the signal and had arrived at certain times that he 
regarded as satisfactory for the initial build-up. 

170 E. W. KELLOGG [j. s. M. P. E. 

Later he spoke of the necessity of accelerating the opening time around zero 
time. Would that create any difficulties such as shutter bump? 

MR. KELLOGG: The tests I reported were not made on an actual ground- 
noise reduction system. They were made by means of a moving mask that un- 
covered an illuminated slit. So far as we could tell, there is no objection to having 
the light begin instantly to change at the maximum rate. 

MR. ALBIN: Regarding the use of the half -wave vs. the full-wave, do you have 
any preference, in view of conditions that exist for example, asymmetry of the 
wave, and the fact that the other half of the rectifier ordinarily would not serve 
any purpose because one-half of the cycle would open the modulator, in any 
event, opposite from the direction that would cause overload? 

In other words, take an analogy to a biased light-valve modulator: One-half 
of the cycle will open the valve, so that that half of the modulation would not 
need a rectifier to operate it. The advantage of the full-wave rectifier then would 
be to utilize the other half -wave and thereby give the noise-reduction cancellation 
a little advance start, so as to gain a little in time. 

Does that again warrant the use of the full-wave rectifier over the half-wave 
rectifier in your opinion? 

MR. KELLOGG: An ideal system seems to me to be one that utilizes full- wave 
rectification but employs some means for preventing peaks that are away from 
the shutter from increasing the opening much beyond that given by the half- 
waves on the shutter side. This, and the possibility of permitting the positive 
waves to contribute only during the start are discussed in the paper. 

As is shown in the illustration of the cathode-ray oscillograms, quite often the 
positive waves appear in appreciable amplitude considerably before the negative 
waves. It would be distinctly desirable to use them, and I think full-wave 
rectification has that to be said in its favor. 

Of course, we all grant that full-wave rectification may reduce the ripple that 
has to be filtered out, but I think that the advantage of this is overestimated. 
Unless there is pretty good symmetry the filtering advantage of the full wave 
tends to be very much cut down. 

MR. ALBIN; The disadvantage of full-wave over half-wave rectification lies 
in the fact that the margin might be excessive if the wave were considerably 
asymmetrical. Is that correct? 

MR. KELLOGG: That expresses it very well, I think. 

MR. ALBIN: You would then say that that is not sufficient to discourage the 
use of the full-wave rectifier; in other words, the faster opening still warrants 
the full-wave rectifier? 

MR. KELLOGG: That is a balance of one thing against another, upon which 
I hesitate to express an opinion. Before I would strongly advocate the half- 
wave against the full-wave rectifier, I would want to make sure that we were 
going as far as possible in taking advantage of polarity, so as to keep the longer, 
higher peaks away from the shutter, and also that we were doing something to 
reduce the ripple at its source, as I have described. With those steps taken 
perhaps the argument would be in favor of half-wave rectification. 

MR. DIMMICK: With reference to Mr. Livadary's question, Mr. Kellogg men- 
tioned the fact that the intensity of the shutter bump depended mostly upon the 
slope, or rate of rise of the unfiltered voltage, and not so much upon shutter 


amplitude. I do not believe he intended to indicate that amplitude is not a 
factor. For a given slope, or rate of rise of the unfiltered voltage, the intensity 
of the sound goes down as the amplitude of the portion having the steep slope 
is lowered. 

In the system to which Mr. Livadary referred, the portion that has the in- 
creased slope is of very small amplitude, and we have found by experience that 
an increased slope can be tolerated over that portion, with a very desirable result. 

MR. KELLOGG: I might add that the sort of input-output curve I was showing, 
with the sharp shoulder, represents the rectified voltage before the filtering. 
After it has been through one stage of resistance-capacity filtering the contribu- 
tion of that little shoulder to increasing the slope of the filtered voltage is almost 
negligible, provided, as Mr. Dimmick has said, that the steep portion is short, 
but it does improve the promptness with which the motion starts. 

MR. LORANCE: The statement was made, I believe, that the problems of 
noise-reduction in variable-area and variable-density recording were quite similar. 

MR. KELLOGG: There are, of course, qualifications to such a statement. The 
mechanical part of it, namely, getting the shutter out of the way, is, I think, 
essentially the same kind of problem in both systems. 

MR. LORANCE: It is. Yet I feel that it is essential to recognize the minor 
differences. We probably have recognized them in the past but perhaps may 
now be overlooking them. For instance, in variable-area recording, to get 10-db 
noise -reduction we must bias to one-tenth the normal width. In variable-den- 
sity recording, to get 10-db noise-reduction we must bias to about three-tenths 
the normal exposure. This means that only about 10 per cent of the total mod- 
ulation can be accommodated in a variable-area system before clipping begins; 
while in a variable-density system with 10-db noise reduction, about 30 per 
cent of the total modulation can be accommodated before clipping begins. 

Filtering probably resolves itself to much the same thing in both systems, 
although the history is such that we may be inclined to overlook the fact. 
A great deal of variable-density recording has been done with the light-valve. 
All the filtering must be in the noise-reduction circuits, because the ribbons 
respond to such a wide frequency range. But in variable-area recording the 
shutter mechanism probably introduces a noticeable amount of filtering. When 
we discuss filtering we should take into consideration the overall combination 
of the noise-reduction device itself, as well as the circuits which feed it. 

MR. KELLOGG: I do not disagree with anything that Mr. Lorance has said. 
If I had had time, I should have elaborated on that subject. I hope I have 
made it clear hi the paper that I am thinking of the ideal system as one that, 
if it does use a shutter, will use one with minimum inertia. I should like to 
reduce the filtering to one stage, and not have a second stage occurring in the 
shutter inertia. Actually, at present, shutter inertia plays a small part of the 
total timing. I look forward to the availability of still faster shutters, and I 
think that it will be possible, regardless of the fast shutters, so to reduce the 
ripple at its source that we shall not need the filtering we now get from the 
shutter inertia. 



Summary. The duplication of 16-mm films involves many relatively intricate 
problems not encountered in the laboratory processing of 35-mm sound-films. These 
problems have given rise to procedures and apparatus radically different from those in 
use in 35-mm. 

The two major differences that are especially significant are (1~) the use of reversal 
for original films; (2) the existence of but one row of sprocket-holes on the 16-mm 

It is interesting to note that all our present standards in 16-mm blindly assume the 
negative-positive method of operation, ignoring entirely the reversal and Kodachrome. 
At the present time even the emulsion position of the 16-mm film is standardized on 
the basis of a 35-mm sound negative and 35-mm picture negative as originals. As a 
result, our 16-mm dimensions so derived from 35-mm are inconsistent with the pro- 
tector dimensions at present in use, and inconsistent with the pressing needs arising 
from the direct 16-mm field. 

Much of the difficulty arises from the rather obvious lack of concern displayed by the 
35-mm entertainment industry and the very rapid simultaneous growth of direct 16- 
mm in educational and industrial applications especially in connection with the dupli- 
cation of sound on Kodachrome. 

Some of the special processes and special apparatus features involved are described 
which have made possible workable solutions to the problems involved. 

It is the avowed purpose of standardization to crystalize practice 
in order to make for better uniformity and greater reliability of prod- 
uct. This has been no less true of 16-mm motion picture practice 
than it has in any of the broader fields of application of technology. 

It has been said that the most standard thing in the world is a 
piece of 35-mm motion picture film; this film will perform in any 
35-mm projector regardless of where either film or projector may be 
made. To a slightly lesser degree the same may be said to be true 
of 16-mm film and 16-mm projectors. Despite the fact that the 
differences between the two cases may be small, they are extremely 
important, and they will explain why, in many cases, prints made in 
35-mm in accordance with 35-mm practices may accomplish what is 

* Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received October 
20, 1940. 

** Precision Laboratories, New York, N. Y. 


desired, while prints made in 16-mm in accordance with 35-mm prac- 
tices miserably fail. There must be an understanding of the medium, 
its advantages, and its limitations before best use can be made of it. 
One of the purposes of this paper is to point out what some of the 
differences are between 35-mm practice and 16-mm practice in the 
production of 16-mm films, how these differences arise, and how the 
problems which result are solved in a practicable manner. 

It may seem trite to state the postulate that a 16-mm film should 
perform satisfactorily in a 16-mm projector. When this statement is 
rewritten in different form, it acquires much significance. We may 
say that, when a 16-mm film made in accordance with what is con- 
sidered good technical practice does not perform with maximum 
satisfaction upon a projector made in accordance with current stand- 
ards and good technical practice, something is wrong with the prac- 
tice, with the standards, or with both. 

Our Society has established recommended practices and standards 
in connection with 16-mm film and equipment, and these practices 
and standards have been discussed in some detail in previous papers. 1 
"A Criticism of the Proposed Standard for 16-Mm Sound-Film" was 
an attempt to compromise in a practicable way among the past, 
present, and future. It has been used as a guide in making decisions 
on how the problems, that arise may be solved in a logical and prac- 
ticable manner. 

Let us first look at a piece of 35-mm film and a piece of 16-mm film 
and compare them. The 35-mm film is more than twice as wide as the 
16-mm. The 35-mm sound-film (and the silent as well) has eight 
sprocket-holes per frame, four on each side; the 16-mm sound-film 
has only one. 

Let us next look at the projectors in which these films are run. In 
the 35-mm projector, the emulsion of the film faces the light, while 
in the 16-mm projector, the emulsion faces the screen and is away 
from the light. While we are concerned with the light we might 
as well find out what quality of light is used in the 35-mm projector 
the light-source is usually an arc with much blue light, while in the 
16-mm projector, the light-source is usually an incandescent lamp 
which is much more yellow. We might also mention while we are dis- 
cussing the projector in relation to the film that it is not at all unusual 
for 16-mm projectors to be operated with appreciably higher magni- 
fication than 35-mm projectors. One case comes to mind in which 
both 35-mm and 16-mm pictures are projected on the same screen in 

174 W. H. OFFENHAUSER, JR. tf. S. M. P. E. 

a theater the 16-mm magnification is more than twice as great in 
this case. For our purpose here, the matter of film speed of 90 feet 
per minute for 35-mm and 36 feet per minute for 16-mm is not of 
great importance. 

Let us now return to a practical production and laboratory problem : 
a 16-mm film is needed that is to be part black-and-white and part 
Kodachrome, and, of course, the film is to be made with sound. 

Taking the black-and white part of the problem first, we would say, 
offhand, that it should not be difficult; all that is necessary is to take 
our picture on 35-mm and everything will turn out all right. We 
take our picture and record our sound and then send the film to the 
laboratory. At this point the problem begins; the laboratory, usually 
without special instructions, observes that the film is 35-mm, de- 
velops it accordingly, and returns it. Picture is ordinarily developed 
to a gamma of about 0.65 to 0.75, since when such a negative is con- 
tact-printed to a gamma of 2.4 or so, the resulting picture is considered 
pleasing when projected on a 35-mm projector. If the sound is 
linear-recorded variable-density, the negative is developed to a 
gamma of about 0.45 or so, since when it is so developed it produces 
the proper overall print gamma of about unity when 35-mm contact- 
printed. It has not been at all uncommon for both picture negative 
gamma and track negative gamma to exceed these "average" figures 

When the negatives are edited, they are then turned over to a 
laboratory for printing (often by a technically uninitiated client) and 
the job of attempting to make good release prints begins. Everything 
seems in good order; the emulsion position of the prints will be in ac- 
cordance with the standard, and the negatives are made in accordance 
with good commercial practice; yet the prints will be washed out and 
hard, and the sound unnecessarily badly distorted. A rather well 
known important fact has been ignored: an optical printer increases 
contrast, and nowhere has this been taken into account. The client 
has been led to believe that all 16-mm prints are bad, and since in 
most cases he does not know the difference and has never been 
shown, he continues to be satisfied with print after print of horse- 
and-buggy quality. At these laboratories, we try to compensate for 
this by obtaining positive film of lower gamma for reduction printing 
than that used for contact printing. In one extreme case we have 
had to go so far as to develop positive prints in a negative bath! 

The trend today is in the direction of high-quality fine-grain films; 


the usual positive emulsions of this type are of slightly higher gamma 
(or contrast) than regular positive emulsions. Thus if the present 
35-mm practice is continued and 16-mm fine-grain high-quality 
higher-contrast films come into wider use, as they are now, our im- 
proved fine-grain print will not improve results to the great degree 
possible and desirable without some material alteration of the han- 
dling of the original 35-mm negatives. In the long run it will be far 
less expensive to reduce the original negative gammas and thus 
allow for the reduction printer and film contrast increase than to 
print all the 16-mm prints on some such low-gamma stock as a dupli- 
cating type or to make some other special provision more costly than 
regular print development. 

At this point, another suggestion may be in order in connection with 
reduction prints from 35-mm with variable- density tracks. In en- 
tirely too many cases, there is no rhyme or reason in the relationship 
of the picture negative gamma to the track negative gamma. It must 
be remembered that both are to be used in making the combined print 
and that when a combined print is made the picture and the track 
must go through the same developer. The right hand must know 
what the left hand is doing. 

The case of the 35-mm picture negative in combination with the 
35-mm variable-area sound-track raises other questions, not the 
least important of which is the fact that most reduction track printers 
are of the fixed reduction-ratio type, and many are not too well ad- 
justed, and therefore the 16-mm prints must be carefully watched in 
the case of even the good 76-mil 35-mm sound-tracks to make cer- 
tain that they will satisfactorily project on a 16-mm projector. The 
width of the track is 76 mils; the reduction ratio is fixed at about 
7 to 6; and from there on, let the chips fall where they may and 
they do. 

A short while ago, one of our customers brought in a 16-mm re- 
duction print made in one of the large New York 35-mm laboratories 
and wanted to know why it was "sour." The reason, upon investiga- 
tion, was quite simple the sound-track on the 16-mm print was 
0.013 inch out of position. In our Standards Committee we are 
much worried about placement errors in printing of 0.002 inch. On 
the whole, West Coast laboratories are less concerned with 16-mm 
prints, with the result that comparable cases are not less uncommon. 

Let us again return to our hypothetical 16-mm film and find further 
alternatives for our black-and-white section. We can photograph on 

176 W. H. OFFENHAUSER, JR. [j. S. M. P. E. 

16-mm negative record our sound as a 16-mm negative and make 
a combined 16-mm contact print of our picture. Good quality is 
possible; there are no optical printing steps to accentuate scratch 
and cinch marks and to increase the harshness of the picture. We 
have, however, one matter that we must watch in connection with 
our sound record it must be made to run in the same manner as the 
picture negative ; otherwise the sound-track, if contact printed, must 
either be printed backward or through the base. In the first case of 
error the result is useless and in the second case the result is noisy 
and almost unintelligible. In neither case of error is it commercial. 
Rule No. 1 in purchasing a 16-mm sound recorder should therefore be : 
use a machine that will run equally well in either direction without any 
mechanical or other changes necessary to effect the change-over from 
operation in one direction to operation in the other direction. 

If our direct 16-mm negative is contact printed, the emulsion posi- 
tion of the 16-mm print is similar to the emulsion position of a 
35-mm print; it would be correct for 35-mm and "wrong" for 16-mm. 

Our 16-mm black-and-white film may be made with a 16-mm re- 
versal original from which a duplicate negative is made, which dupe 
negative, together with a sound negative, is used to make a contact 
print. Both picture quality and sound quality in a print from such 
a combination may be made superior to picture and sound quality 
obtained by any other means. Some data concerning the picture 
phase of this method have been published in the JOURNAL by J. A. 
Maurer. 2 In this case the emulsion position is correct (assuming 
contact printing throughout). By optical printing from the reversal 
to dupe negative, the emulsion position may be made "wrong." 

Let us now go to the Kodachrome section of our hypothetical reel. 
Kodachrome is a reversal film and therefore the emulsion position is 
"standard" away from the projection lamp and toward the screen. 
If we want a Kodachrome duplicate, it is best made by contact 
printing rather than by optical printing; optical printing would not 
only further increase the Kodachrome contrast but would also greatly 
accentuate all the scratch, abrasion, and cinch marks on the original 
film. When Kodachrome is contact-printed, the emulsion position, 
as in the case of negative, is "wrong" in projection the emulsion 
faces the light; also in the case of negative-positive on 16-mm the 
black-and-white print is also "wrong." The two may then be spliced 
together and when they are run, they will project without change in 
focus of either picture or sound-track, throughout the reel. 


Another practical problem that arises relates to the printer aper- 
ture for Kodachrome. Kodachrome is essentially a reversal film and 
a Kodachrome duplicate is therefore equivalent, in a sense, to a dupli- 
cate reversal. If a duplicate is to be printed, it is obvious that the 
picture printer aperture should be smaller than the camera aperture 
if a white frame line around the picture is to be avoided. The pro- 
jector aperture should be smaller than either the printer aperture or 
the camera aperture if both black or white frame lines are to be 
avoided in projection. The sizes of these apertures must be carefully 
studied, as it is not unusual for 16-mm sound projectors to have a 
picture aperture far smaller than standard which, with composition 
intended for standard aperture projectors, results in the rather an- 
noying cutting off of heads and feet. With apertures related in this 
manner instead of in the customary negative-positive manner, appre- 
ciably greater camera, printer, and projector accuracy is required. 

The next matter for consideration is the sound. We can use either 
a 16-mm original or a 35-mm original, either variable-density or 
variable-area. In 16-mm the originals may either be developed and 
printed or an original may be recorded which is reversed either chemi- 
cally or optically quite a variety of types when we compare this 
with 35-mm. 

We can set down several general rules for printing sound on Koda- 

(1} The printing aperture should be smaller than the positive image from 
which printing is accomplished. 

(2) If a variable-area negative is used, the printing aperture should be larger 
than the width of the negative record. 

(5) The projector aperture should be wider than the 16-mm recorder aperture 
(or reduced 35-mm aperture) ; yet smaller than the aperture used for printing the 
black-and-white positive. 

Several comments should be made at this point concerning vari- 
able-density linear recordings. With regard to the quality of sound, 
a good general rule of thumb is to duplicate from a positive which in 
itself has minimum distortion. In the case of variable-density 35- 
mm originals, it should be obvious that a 35-mm negative of gamma 
0.45 which will contact-print in the usual way to a gamma of 2.4 will 
have an excessive gamma product when optically reduced from that 
35-mm positive to Kodachrome ; the increase in printer factor of the 
optical sound-printer over the contact printer is one possible cause. 
Our experience indicates that there are others, as we have found that 

178 W. H. OFFENHAUSER, JR. [j. s. M. P. E, 

a special low-gamma 16-mm reduction-track positive seems to cor- 
rect not only the JKodachrome duplicating step but also for other 
steps as well. The track positive gamma is so low that the distortion 
produced by a print of conventional gamma is best described in some 
cases as "terrific." 

The case for 35-mm variable-area negative tracks has the usual 
complications with regard to dimensions and envelope effect. If the 
35-mm negative is optically reduced to a 16-mm positive, as we do at 
this Laboratory, and this 16-mm print used for the Kodachrome du- 
plicate (as in the variable-density case above), only the 16-mm track 
printer needs to be masked to print to Kodachrome from any conven- 
tional original negative either 16-mm or 35-mm. At this point, we 
must again warn against the not uncommon uncertain sound-track, 
the uncertain 7 to 6 reduction ratio, and its uncertain result. 

The optical reduction of 35-mm sound negatives to 16-mm smooths 
one other difficult problem: that of synchronizing the picture and 
sound. In 35-mm the sound is advanced 20 frames; in 16-mm the 
sound is advanced 26 frames. At our laboratories all 16-mm films 
are marked even (this is easier for cutting), and the advancing is done 
in the printing room on the printing machine. In this manner we 
avoid synchronizing-mark confusion, since the question of whether 
films are advanced 20 frames as in 35-mm or 26 frames as in 16-mm 
does not arise. 

With regard to other sorts of originals for sound such as toe-re- 
corded 16-mm tracks or toe-recorded 35-mm variable- density tracks, 
it can be said that they are very few in number, and for the most 
part the quality is quite poor due to the fact that the Kodachrome 
duplicating step is usually ignored, and for other reasons not related 
to Kodachrome. Variable-area master positives in either 35-mm or 
16-mm have not put in an appearance; some misguided attempts by 
the technically misinformed have been made to save the cost of a 
track positive, but the sound proves so poor as a result, and the pro- 
cedure so costly in the long run, that such attempts die out very 
quickly, usually after the first attempt. 

Let us return again to our production and laboratory problem: 
the film, part black-and-white and part Kodachrome. If we want 
acceptable picture quality, we are forced to contact-print the Koda- 
chrome. The sound is no problem; we print our sound on a 16-mm 
optical 1 to 1 sound-printer. With this printer it is possible to focus 
on either side of the film merely by turning a focusing ring as far as it 


will go in the proper direction. Our combined duplicate, though good 
in quality, is "wrong" ; it is non-standard. 

If we are to splice black-and-white together with Kodachrome in 
the same reel, it is necessary to make our black-and-white "wrong" 
or non-standard as well; in this manner we avoid the necessity of 
refocusing our projector in the middle of the reel. Progressive pro- 
jector manufacturers long ago recognized the importance of "wrong" 
or non-standard films, and made provision for focusing the sound 
optics of the projector on either side of the film. Thus with a ' 'wrong' ' 
projector and a "wrong" film, we obtain best results with our prob- 
lem picture. 

If we return now to our premise that the desideratum is that a 
16-mm film projects best in a 16-mm projector, it would seem, in this 
case at least, that the avowed purpose of our 16-mm standard is de- 
feated under the present standard. Without qualification, it is 
strongly urged that the 16-mm standard for the emulsion position be 
made to conform to the 35-mm standard in projection that is, 
toward the light. It is also recommended that, if the 16-mm standard 
is to remain tied to 35-mm, some analysis of the 35-mm problem in 
relation to 16-mm be made by those interested in the 35-mm pro- 
duction of 16-mm films in the light of the aforementioned problems. 
In the absence of such an analysis and recommendation and in con- 
sideration of the complexity of the 16-mm problem, a complete di- 
vorce of 16-mm from 35-mm is recommended via a specific optical 
reduction ratio or other proportional dimension relationship. It is 
only in this manner that we can right a growing "wrong." 


1 MAURER, J. A., AND OFFENHAUSER, W. H.: "A Criticism of the Proposed 
Standards for 16-Mm Sound-Film," /. Soc. Mot. Pict. Eng., XTCX? (July, 1938), 
p. 3. 

FRIEDL, G.: "Some Data Regarding Dimensions of the Picture Image in 
16-Mm Reduction Printing," /. Soc. Mot. Pict. Eng., XXVIII (June, 1937), p. 

MAY, R. P.: "Sixteen-Mm Sound-Film Dimensions," /. Soc. Mot. Pict. Eng., 
XVHI (April, 1932), p. 488. 

2 MAURER, J. A.: "Commercial Motion Picture Production with 16-Mm 
Equipment," /. Soc. Mot. Pict. Eng., XXXV (Nov., 1940), p. 437. 



Summary. The March of Time requires equipment of great portability and sim- 
plicity of operation, yet retaining good quality. By using Class B push-pull, vari- 
able-area recording, a complete noise-reduction sound system weighing fifty pounds 
was obtained. This single system was used in production of the feature picture The 
Ramparts We Watch. Problems arise from (1} recording on panchromatic negative, 
(2} lack of control over negative processing, (5) instability of recording unit caused 
by rough use of camera on which it is mounted, and (4) distortion due to lateral track 
shift. Means for overcoming these handicaps sufficiently have been found. Single 
system can be used without great sacrifice in quality, where time and space are factors. 

The nature of March of Time work requires light-weight equipment 
of extreme portability and simplicity of operation. The equipment 
must be such that it can be moved into position and set up for a shot 
in ten minutes or less. Most scenes are made on location in offices 
of government officials, in coal mines, in aeroplanes, in battleship 
turrets, at political meetings; in short, on typical newsreel locations, 
where both time and working space are limited. However, better 
than usual newsreel quality of sound is essential. 

It is reasonably simple to construct a light-weight amplifier with a 
good response characteristic. To retain some degree of versatility 
and still keep bulk and weight at a minimum, a novel but simple 
input layout was devised in the amplifier design. A two-position 
mixer is used with a three-microphone input. One input line feeds 
directly through one mixer, while the other two lines with a selector 
switch between them feed the second mixer. As a result an amplifier 
of small dimensions was obtained, having varied capabilities but 
weighing only forty pounds, complete with batteries sufficient for 
twenty thousand feet of recording. Using small cables and light- 
weight microphones of the highest quality available, the entire equip- 
ment with the exception of the recording unit obviously can be made 

* Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received October 
12, 1940. 

** March of Time, New York, N. Y. 



adequately portable. For the March of Time, the weight and bulk 
of a separate recorder were out of the question. Also, any heavy, 
complex, single-system camera was equally unsuitable, as was any 
additional equipment for noise reduction. Hence a standard newsreel 
camera, recording on the main sprocket, was selected. 

Request was made for the development by RCA of an optical sys- 
tem to give an ultraviolet, Class B, noise-reduction sound-track. 
It was required that this unit be compact and light in weight, and 
adapted to the newsreel camera selected. To aid in keeping the unit 
compact, the frequency response upper specification was reduced to 
6500 cps. It was believed that a loss at frequencies above 6500 would 
not noticeably affect speech quality and would not render any field 
music recording unusable. The galvanometer unit, more than meet- 
ing these requirements, adds little bulk to the camera and approxi- 
mately five pounds to its weight. Thus, without the necessity for 
additional noise-reduction equipment, or for a double-system recorder, 
a small, complete, good-quality sound system, weighing approxi- 
mately fifty pounds, was obtained. The entire sound outfit is carried 
and operated by one man, eliminating the necessity for a large truck 
or a recording room, with the accompanying cables and intercommuni- 
cating systems. 

For the feature picture The Ramparts We Watch, the same re- 
quirements of lightness and portability of equipment existed. The 
producers desired to inject an authenticity and feeling of reality into 
their scenes by working on actual locations so far as possible, and to 
eliminate unnecessary expenditure for sets and equipment in achiev- 
ing that reality. The troublesome points in the single system had been 
sufficiently ironed out that it was believed a satisfactory job of re- 
cording the first feature-length picture could be done. 

The principal difficulties with the system are (1) recording on 
panchromatic negative stock, (2} lack of any control over processing 
the negative sound-track, (3) a tendency toward instability of the 
sound unit on the camera, and (4) susceptibility of the push-pull 
track to distortion due to lateral shift in track position. Means have 
been found for adequately overcoming these troubles. 

In the first place, variable-area recording on negative stock means 
reduction of signal-to-noise ratio, because it is obvious that exposure 
of the negative track can not be made very great before lack of defi- 
nition becomes serious. However, it was found that with Class B 
track, this situation could be sufficiently overcome. By exposing 

182 D. Y. BRADSHAW tf. s. M. P. E. 

the track to give a negative density of 1.2 to 1.3 when the film is 
given normal picture development, and printing to a density of 1.3, 
good contrast is obtained with only slight loss at high frequencies. 
This is due in great part to an inherent characteristic of the Class B, 
variable-area track. Each half of the wave is recorded on an en- 
tirely independent track, so that both the maximum and minimum 
points on the original wave are exposed on the negative at points on a 
convex curve, with a half-cycle length of unmodulated track separat- 
ing the adjacent exposed areas. In the case of other variable-area 
tracks, the minimum points occur where the exposed section is con- 
cave, with no separation between adjacent exposed areas. Although 
the filling in is reduced to a great extent with a double-system 
set-up, using positive sound stock, the loss of definition when recording 
on negative stock is objectionable. However, the separation obtained 
with Class B maintains the high-frequency response to a surprising 

Because of limited facilities and time, no attempt is made to check 
the negative gamma. Even if the negative gamma were known, the 
wide variations in negative picture development would make useless 
any attempt to obtain overall gamma of unity. However, the effect 
of variation in processing can be greatly offset. Cases of extreme 
overdevelopment have been encountered, but have been corrected 
by an increase in print density. The same treatment has been used 
to correct for overexposed tracks. The worst case occurred when a 
cameraman switched to a faster film at the last minute before photo- 
graphing and recording Fred Waring' s orchestra for a commercial 
release. The soundman was not notified of the change and exposed 
the track for the usual panchromatic film. When a print of the track 
of density 1.2 was reproduced, the distortion was intolerable. How- 
ever, by printing to a density of 1.5, the roughness disappeared from 
the track, and the resultant increased attenuation at high frequencies 
was not greatly objectionable. 

When the tracks are underdeveloped, the prints are made slightly 
on the light side and good results are obtained with some sacrifice of 
signal-to-noise ratio. The worst condition encountered involved a 
negative track density of 0.7, from which a print of density 1.05 was 
made. This particular recording was used in one of the scenes of 
The Ramparts We Watch and was not objectionable. 

A newsreel camera is not the ideal location for a sound modulator. 
Packed and unpacked several times a day, transported over the 


country in the trunk of a camera car, carried over the cameraman's 
shoulder, operated at all angles the rough use of the camera results 
in abuse of the galvanometer unit as well as of the sprockets and guide 
rollers in the camera. Nevertheless, a simple correction has been 
found for each difficulty encountered. Adjustments usually can be 
made after inspection of the equipment on location, but on occasion 
equipment failure has produced unique problems. In one instance, 
the lamp filament sagged in the middle of a long sequence. The re- 
sult was a normally exposed negative on one side of the push-pull 
track while the other part was badly underexposed. Densities were 
1.2 and 0.7. The speech from the reproduced print was found to be 
unintelligible. However, a simple adjustment of the balance of 
photoelectric cell voltages on the reproducer compensated for the 
unbalanced track and a satisfactory reproduction resulted. 

Use of a newsreel camera without adequate provisions for insuring 
a fixed lateral position for the film results in a variation from scene to 
scene of as much as 5 mils. An attempt to align the film with rollers 
resulted in buckling the film as it left the intermittent sprocket. As 
the present problem is much more pronounced in printer variation, the 
purely mechanical solution of film position in the camera has not been 
attempted as yet. In an extreme case, a properly aligned negative 
track has been printed fifteen mils out of position, though the error, 
when present, does not generally exceed ten mils. The March of 
Time staff works to a deadline, without exception, and time seldom 
exists for reprinting displaced tracks. Obviously, a push-pull track 
is more susceptible to distortion from this cause than is a standard 
track. It has been found that the displacement must be enough to 
take at least two mils off the maximum peaks, to give noticeable 
speech distortion. Only one side of each half of the track is affected, 
of course. To overcome the situation in practice, some concessions 
are necessary. The recording is made at a maximum peak modula- 
tion of 80 per cent instead of 100 per cent. This means a latitude in 
track position of four mils on each side of each half-track. Hence the 
camera and printer must introduce an error of more than six mils to 
cause appreciable square-topping of the wave. The separation be- 
tween the two parts of the Class B track is six mils. Any error in 
excess of ten mils becomes exceedingly serious, as the half-cycle 
peaks on one part of the track are then scanned on the other part, 180 
degrees out of phase. When such a case exists, and no time is con- 
veniently available for a reprint, it is a simple matter to move the 


push-pull head on the reproducer an amount equal to the displace- 
ment. By setting the scanning so as barely to miss the sprockets, a 
satisfactory re-recording has been made from the extreme case of 
fifteen mils' displacement. Use of a dependable printer and camera 
will eliminate the problem of lateral shift entirely. 

In ninety per cent of March of Time work, no trouble has been 
encountered and no corrections have been necessary. However, the 
single-system unit is susceptible to these difficulties which have been 
adequately overcome when they have appeared. Application of these 
occasional corrections is the only penalty for the use of a very port- 
able, inexpensive, easily operated sound system. It is not suggested 
that the industry discard the comparatively cumbersome and expen- 
sive double-system equipment and the additional personnel required 
to operate and maintain it. Certainly such equipment is necessary 
to obtain the present limit of high-quality sound. Where time and 
room are factors, however, the single system can be used without 
great sacrifice in quality. 



Summary. Television utilizes certain elements of operative photography. In 
live-subject presentations these include composition, focus, contrast range, intra- 
image contrast of one object from another, dolly shots, panning, and certain aspects of 

In television, operative maneuvers must be quickly and smoothly executed. The 
camera in question may be supplying the outgoing image at the time in question, or, if 
not, it should rapidly be made available for change in camera angle on the program. 

The equipment and technic evolved at W6XAO to meet these requirements during 
several years of telecasting are described. 

Television utilizes certain elements of operative photography. 
The television camera is maneuvered like a motion picture camera, 
only more intensively so. In studio cinematography opportunity is 
afforded to pause between sequences in the action, to check camera 
angles, composition, focus, lighting, and to predetermine the next pan 
or dolly shot. In television the action is continuous. Although two 
or three cameras simultaneously televise a scene, each from a differ- 
ent angle, it is necessary that each follow the action most of the time, 
thereby to be available to carry the outgoing picture at the election 
of the television producer. In this respect the cameraman's function 
is similar to that in newsreel cinematography, where the action must 
be followed and composition, focus, and other operative adjustments 
maintained almost subconsciously. 

In considering apparatus, there are two major differences in con- 
struction and operation depending upon whether the focusing is ad- 
justed by the cameraman or remotely adjusted by a television engi- 
neer who views a monitor image. There are three minor differences, 
relating to the type of viewfinder, whether it is a duplicate of the 
camera tube optical system, whether the camera tube image is viewed 
by mirror reflection, or whether a separate viewfinder is utilized, 

* Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received October 
1, 1940. 

** Don Lee Broadcasting System, Los Angeles, Calif. 


186 H. R. LUBCKE (J. S. M. P.E. 

either of the cinematographic type or of the frame and peephole type. 

In operations it can be said that each of the above possesses ad- 
vantages and disadvantages. 

It can also be said that a trained crew can produce satisfactory 
television programs with any of the apparatus mentioned. The opera- 
tive cameraman is always provided with head-phones and informa- 
tion transmitted thereover serves to bridge the shortcomings of either 
apparatus or personnel. Specifically, the cameraman may be warned 
of an impending change in location of the action, or instructed to fol- 
low this character while allowing another to pass from his field of 
view. Faults in composition or other image characteristics are at 
once relayed to him. 

Rehearsal is the "great polisher" which contributes to the smooth- 
ness of the show. Rehearsals increase the expense of operations, how- 
ever, and in days of non-commercial (non-income) television a fine bal- 
ance is maintained between rehearsal time, skill of the crew, and the 
quality of the result. At times, a happy combination indicates the 
perfection of future television productions; at other times, ines- 
capable accidents give a momentary result which cries for that saving 
grace of cinematography, the "retake," although this is impossible 
for obvious reasons. One faculty created by a skilled crew is the 
ability to gloss over errors or mistakes. An attentive producer elimi- 
nates faulty camera operation by shifting instantaneously to an- 
other camera. A fade covers a multitude of sins. It is an inspiration 
to watch emergency departures from the script skillfully handled by a 
production crew, resulting in a smooth performance to all save the 
few who knew what should have been. 

This emphasis on the personnel aspect of television operations has 
been impressed upon us at W6XAO by years of daily telecasts. Ap- 
paratus is only part. The welding of apparatus and personnel into 
an operative whole is the important thing. Precision gained through 
apparatus complexity from an operative standpoint is to be decried 
rather than to be sought after. 

To return to the apparatus and the major differences in construc- 
tion; there are advantages to remote focusing. The cameraman is 
relieved of a duty and the focus adjustment is made utilizing an indi- 
cator which counts most, the actual television image. However, 
careful attention to focusing must be given by the engineer or the 
action will "get ahead of him." Also, such focusing requires one more 
man in the video booth, increasing personnel costs. 


The duplicate optical system viewfinder is usually used with cam- 
eraman focusing and is quite satisfactory, only vertical parallax and 
cost and weight of the camera being adverse factors. 

The mirror- viewing camera-tube-image viewfinder has no parallax, 
but tends to give an image too dim for consistent accuracy of focusing. 
Present technic uses an aperture from//3.5 to//8. It is obviously ac- 

FIG. 1. The mirror- vie wing camera-tube-image viewfinder. 
The cameraman is viewing the image on the photoelectric 
plate, with right hand on the focusing adjustment on the 
farther side of the camera. The left hand is used for panning 
and tilting as shown. The scene is a close-up. The exagger- 
ated under-eye make-up is no longer used. 

curate in determining composition, even on close-ups, where the 
separate type of finder encounters difficulty. A camera of this type, 
with cameraman focusing, is shown in Fig. 1. 

The usual cinematographic viewfinder has been tried in our opera- 
tions but found unsuitable unless provided with automatic parallax 

The peep-hole and wire-frame type of viewfinder has been used con- 
siderably with remote focusing. This has proved reasonably satis- 

188 H. R. LUBCKE tf. S. M. P. E. 

factory with an adjustment technic as follows. Prior to the perform- 
ance the cameraman checks the field of view of his finder with that 
actually appearing on the monitor image of the engineer. This is 
aligned by adjusting the sides of the finder frame until substantial 
coincidence is secured. Two pairs of side markers are used in the 
type of finder which is located at the side of the camera. The 
markers delineating a line of sight more nearly parallel to the camera 
tube optical axis is used for long shots, and the markers converging 

FIG. 2. Peep-hole and frame type viewfinder. The photo 
shows a spring type adapted to work against a cam on the lens 
housing for automatic parallax correction. The cameraman 
stands closer to the peep-hole in usual operation. 

toward the axis for close-ups. The shift from one group to the other 
is made subconsciously by the cameraman after a few hours' training. 
It is possible to include automatic parallax correction in a finder of 
this type, a cam operating against the motion of the lens mount 
automatically shifting the side markers, as shown in Fig. 2. These 
arrangements have largely been developed by our chief cameraman, 
Dwight Warren. 

Other aspects of operative camera work include dolly shots and 
panning. These are usually indicated on the script and are univer- 
sally called for by the producer viewing the outgoing image. The 


camera crew carries out the instructions given to both the operative 
cameraman and the studio producer. Experience largely determines 
the necessary extent of these movements, and the result is guided to 
satisfaction by auxiliary commands from the producer to the effect 
of "more" or "hold that." as the case may be. 

Still other aspects of television "photography" comprise contrast 
range, intra-image contrast of one object from another, and the effect 
of lighting, make-up, costumes, and properties. 

The contrast of the image is adjustable electrically within certain 
optical limits. The lighting must be sufficient to produce optimum 
image brightness on the photoelectric surface of the camera tube. 
If an excess is available, as in outdoor pick-ups in sunlight, the con- 
dition is easily and happily remedied by closing the aperture. This 
gives relatively great depth of focus and often a discernible increase 
in image quality because of improved lens performance. Within the 
optical limits, camera tube beam current and amplifier gain deter- 
mine the contrast. This is usually adjusted by the television engi- 
neer to "fill the pedestal"; that is, to utilize the full electrical wave- 
form characteristic from the value representing black to the value 
representing full white; but not further, which will result in "cutting 
off the whites," "cutting off the blacks," or both, and is equivalent 
to over or underexposure in photography. 

Intra-image contrast is a matter to be determined by rehearsal 
tests or to be provided for by prior experience. Almost any scene will 
televise intelligibly; but as in photography, the striking symphonies 
of light and shade which characterize the work of the expert must be 
planned in deliberate detail. The essential factor is the tone rendi- 
tion of the colors and shades of everything that is in the scene. The 
color response of the usual camera tube is roughly panchromatic with 
overemphasis in the red. The spectral response of the incident light 
is a factor. At W6XAO we have found that white light of high-tem- 
perature tungsten lamps is preferable to the yellow-red light of ordi- 
nary lamps. Panchromatic make-up with violet-hued lipstick is sat- 
isfactory. "Indian pottery tan" in all its tones appears to be a 
uniquely satisfactory color, as has been previously reported. l Proper- 
ties are easily assembled which prior tests have shown to be satis- 
factory with the studio technic and characteristics in use at a given 

The latter is an important aspect of production operations. It is a 
combination of electrical circuit behavior all the way from camera 

190 H. R. LUBCKE 

tube to average receiver screen, combined with camera optical and 
photoelectric properties in addition to lighting, make-up, costume, 
and property technics. 2 - 8 

Our work has shown that a substantial change in any of these 
factors, at times unintentional, may require compensating changes in 
the technic of one or more of the others. One of the compensations 
of present-day television production is that these changes can be 
made. Another is the preception of the future, with the work of today 
recognized as merely a rough approximation of what can be done in 
the future when greater control will be had over factors now identified 
but not yet susceptible to accurate control. 


1 LUBCKE, H. R.: "An Introduction to Television Production," /. Soc. Mot. 
Pict. Eng., XXXIII (July, 1939), p. 54. 

2 PROTZMAN, A. W.: "Television Studio Technic," /. Soc. Mot. Pict. Eng., 
XXXIII (July, 1939), p. 26. 

3 PORTERFIELD AND REYNOLDS: "We Present Television," W. W. Norton & 
Co. (New York) 1940. 


During the Conventions of the Society, symposiums on new motion picture ap- 
paratus are held, in which various manufacturers of equipment describe and demon- 
strate their new products and developments. Some of this equipment is described 
in the following pages; the remainder will be published in subsequent issues of the 



General film appearance and picture quality have long been major problems 
with the motion picture industry. Of the numerous factors influencing screen 
quality, this paper will deal with two, namely, film abrasion and oil mottle. 

The former of these, film abrasion, has been the object of much research leading 
to a large number of patented processes, some of which are in use in the trade. 
These processes, in general, leave something to be desired, either in effectiveness 
or in price, frequently in both. They depend for their effectiveness on the prin- 
ciple that the treated surfaces will be more resistant to abrasion than those not 
similarly treated. Whereas this may be true in some cases, the fact remains that 
no practical film surface has been found which will resist abrasion indefinitely. 
Therefore, when these treated surfaces become abraded, they present the same 
problem as do any other scratched films. 

Protection of Film from Abrasion. Experiments have been conducted in this 
laboratory and in the field with a new type of film treatment based on a new prin- 
ciple. The aim has been to devise a lacquer which can be applied easily and re- 
moved easily. This lacquer, when applied to both sides of the film, becomes 
scratched just as the film surfaces would have been scratched by any sharp points 
coming in contact with it. If the thickness is correct, however, normal scratches 
do not go through the lacquer layer into the film. Therefore, on removal and 
renewal of the lacquer, the film is found to be in as good condition as when new. 

As for the lacquer itself, it was necessary that it fulfill certain definite require- 
ments. These requirements were: 

(1) Its manner of application must be simple, requiring practically no special 

(2) Its rate of application must be comparable to average processing speeds. 

* Presented at the 1940 Fall Meeting at Hollywood, Calif., October 21-24, 
1940; received November 25, 1940. 

** Eastman Kodak Company, Rochester, N. Y. 




(5) It must be easily removable without the aid of solvents or special equip- 

(4) It should be applicable to both sides of nitrate and safety films without 
any deleterious effect upon physical properties, such as curl, flexibility/moisture 
absorption, etc. 

(5) It must dry rapidly to give a smooth coating of glossy appearance. The 
reason why a glossy surface is essential will be discussed later. 

(6) It should make the films fingerprint -proof . 

Let us now see how the Eastman Protective Film Lacquer fulfills these require- 

Application of Lacquer by Wick Method. Fig. 1 shows the simplest manner by 
which this lacquer may be applied to either 16 or 35-mm films. In this method 

FIG. 1. Apparatus for applying Eastman Protective Film Lacquer to the 
entire surface of 16 or 35-mm film by means of a wick: (A) stock roll; (B) 
drive sprocket; (C) idle roll; (D) plush wick and holder; (E) drying cabi- 
net; (F) take-up roll. 

the entire width of film, perforated area included, is coated with the lacquer. 
This is accomplished by drawing the film from the stock roll over a strip of plush 
wet with the lacquer and thence into the drying cabinet and onto the take-up roll. 
In the case of this plush wick method, the thickness of the coating will depend 
somewhat upon the speed of application. At high speeds the coating may be- 
come quite thin due to the inability of the wick to absorb liquid rapidly enough. 
If the coating becomes too thin the covering power of the lacquer will be impaired 
and proper protection against abrasion will not be provided. It is suggested that 
this method be used only for speeds of application of 50 feet per minute or less. 

Fig. 2 shows a close-up of the wick and lacquer pan. A suitable wick may easily 
be constructed by clipping a strip of plush around a glass rod by means of a wire 

Fig. 3 shows a plush or felt-covered roll which may be used in place of the sta- 
tionary wick. The roll is driven slowly in a direction opposite to that of the film 
travel. In this case the thickness of the coating is independent of the film speed, 

Feb., 1941] 



as the speed of the roller wick may be increased in the same order as 
the film speed is raised. The speed of application, therefore, will be limited 
only by air cabinet capacity, etc. Applications should be satisfactorily made at 
speeds as high as 150 feet per minute. 

Application of Lacquer by Bead Method. Fig. 4 shows the apparatus for apply- 
ing the lacquer to the picture and sound-track area only of 35-mm film. In this 
method the film is wrapped tightly about the idle roll D and the lacquer is applied 
by means of an auxiliary roll E which revolves slowly in a pan of lacquer. This 
applicator roll E does not quite come in contact with the film in operation but at 
the beginning of the coating it is brought momentarily into contact with the film 
on the idle roll D, and then the two are separated slightly. The bead of liquid 
which forms between the film and the applicator roll is thereafter maintained and 
the coating is accomplished by this liquid bead and not by contact with the appli- 
cator roll. By this means a smooth layer of lacquer is applied onto the film from 
perforation to perforation. These edge lines are quite sharp and may be main- 
tained to within about 0.002 inch from a straight line. This method, although 

Plush Wick 
Gloss Rod 

Lacquer Pan 

FIG. 2. Close-up of plush wick and lacquer pan. 

more complicated than the wick method, gives coatings of greater thickness and 
better appearance. It is especially adaptable to negatives, duplicating nega- 
tives, master positives, and the like, since no lacquer reaches the perforated area 
and therefore can not cause trouble from the standpoint of dirt, positioning, etc. 
In the case of this bead method, the thickness of the coating is dependent not only 
on the relative speeds of the film and applicator roll but upon the distance be- 
tween the applicator roll and the film. By this method lacquer has been continu- 
ously applied to 35-mm films at a speed of 150 feet per minute. 

Removal of Lacquer. The protective coating can be removed easily by immer- 
sion for one or two minutes in 1-per cent sodium carbonate at 65 F, followed by a 
clear water rinse. In practice, therefore, the coating can be removed easily and 
completely by passing the film through a machine similar to a processing machine 
hi which the film is given a two-minute immersion in 1-per cent sodium carbonate 
followed by a clear water rinse (Fig. 5). In order to avoid the possibility of emul- 
sion-stripping in this removal process, it is recommended that the sodium carbon- 
ate solution contain 1 per cent by volume of commercial formalin solution. This 
will prevent swelling of the emulsion and likewise minimize the possibility of 
emulsion-stripping on films which have not been hardened sufficiently in the origi- 
nal processing operation. 


Fig. 6 illustrates the removal of the lacquer in a regular processing machine 
merely by the substitution of a neutral rinse for the customary acid stop bath fol- 
lowing the developer. 

Effectiveness of Treatment. Thus the requirements in regard to simplicity of 
operation and speed of operation have been satisfied. Furthermore, comparison 
of coated and uncoated films both immediately after coating and after service in 
the field has shown that there has been no noticeable change in physical proper- 
ties. In addition, fingerprints may be removed completely from the coated film 
by gentle wiping. 

Films treated with this lacquer will be protected against all ordinary cinch 
marks and against the normal scratches found on most films which have been in 
service in the trade. It would be ridiculous, of course, to pretend that any lac- 
quer of a thickness of 0.0001 inch could not be scratched through, if conditions are 
severe enough. Our experience, however, indicates that such scratches seldom 
occur in practice. 

There is one other point in regard to this scratch-protective layer that should 
be mentioned. It has been pointed out that our aim was to apply a lacquer 

Plush or felt wick 

Hard rubber 
OP metol roller. 

'Slot for fasteninq wick. 

FIG. 3. Plush or felt-covered roll for continuous high- 
speed wick application of lacquer. 

which would bear the scratches which would normally be found on the film. 
The question naturally arises, "To what extent does the coating itself become 
scratched? Does it scratch more or less readily than normal film surfaces?" 
This question can be answered at the present time only in the following way. 
Laboratory comparisons have indicated that the coated films have approximately 
the same scratch resistance as untreated films. However, without a single ex- 
ception, the experience with these coated films in the field has indicated that they 
are definitely more resistant to abrasion than the uncoated checks. The ultimate 
answer to this question must be deferred until more practical information has been 

Protection of Film from Oil Mottle. To this point we have been concerned with 
film abrasion. We will now consider the closely allied subject of oil mottle or, in 
other words, the continual flicker on the screen due to oil spots on the film. In the 
course of our study, it soon became apparent that flicker due to oil on the film was 
more detrimental to screen quality than was the occasional scratch. Scratches 
which are extremely prominent to the technical people of the industry nearly al- 
ways go completely unnoticed by the average theater patron, due, no doubt, 
to absorbing interest in the story. On the other hand, flicker on the screen must 
be avoided by all means. Noticed or unnoticed, this mottle most surely has its 

Feb., 1941] 



effect upon the eye and upon the fatigue of the spectator. Although no scientific 
proofs of this are available, I believe that the comments of the spectators who are 
allowed to see both clean film and oily film, one after the other, are sufficient 
indication of the increased pleasure in viewing the mottle-free film. 

This question of oil on the film has not had the attention given to it which it de- 
serves. Heretofore, it has not been thought of as an actual damage to the film as 
is the more conspicuous scratch. Furthermore, one has thought that if oil does 
get on the film it can be removed by cleaning. It is true that oil may be removed 
easily from a small area of film with a clean pad and fresh carbon tetrachloride, 
but it is quite another matter to clean an entire roll effectively without streaks, 
bloom, abrasion marks, etc. Thus it is that oil, which often gets on the film on its 
initial run, regardless of the quality of the house, usually stays there throughout 
the life of the film. Large sums of money are spent by film manufacturers, proc- 
essing laboratories, and studios in order that the photographic quality of the 

FIG. 4. Apparatus for applying protective film lacquer to the picture and 
sound-track area only of 35-mm film by means of a liquid bead: (A) stock 
roll; OB) drive sprocket; (C) float roll (oil-damped); (D} idle roll; () 
driven applicator roll; (F) drive sprocket; (G) drying cabinet; (H) take-up 
roll ; (/) lacquer pan. 

pictures may be maintained at the highest possible level, yet this oil mottle often 
nullifies completely the careful work which has been done on the picture to this 

The reason why oil spots on film produce mottle is well understood. Each oil 
spot produces a glossy surface which permits more of the light from this area to 
be focused on the screen than from the neighboring unoiled surfaces. 

The remedy, of course, is to make the whole surface glossy so that there will 
be no more light coming from the oily spots than from the rest of the surfaces. 
This lacquer accomplishes this to a remarkable degree. Although the trained eye 
can readily distinguish the oil, even on the lacquered film, the improvement is 
great enough so that most spectators would feel that the mottle is entirely elimi- 

Results of Field Tests. For the evaluation of the effectiveness of this lacquer 
treatment under actual trade conditions, a feature picture was placed at our dis- 
posal by the Metro-Goldwyn-Mayer Company. A portion of this print was 
given the lacquer treatment at the time of release. The entire feature was then 



put into service through the Buffalo Exchange of M-G-M. At intervals the print 
was brought to the laboratory for examination. This allowed a comparison to be 
made between the untreated sample and the lacquer-treated sample in respect 



Carbonate -formalin 

Neutral Rinse 

Drqinq Cabinet 

FIG. 5. Apparatus for removal of protective film lacquer by means of carbon- 
ate-formalin solution. 

to abrasion and oil mottle. Screen tests of the two samples clearly indicated 
that the abrasion of the lacquered sample was considerably less than that of the 
untreated sample. Likewise even though there was the same amount of surface 
oil on both films, the mottle on the screen due to this oil on the treated film could 

Stock. Roll 



Take -up 


Wash Drqmq Cabinet 

FIG. 6. Regular processing machine used to remove the protective film 


be detected only with difficulty, whereas that of the untreated sample was very 

When sufficient reduction of screen quality resulting from abrasion was noted 
on the treated samples, portions of them were retreated, i. e., the original lacquer 
removed in carbonate solution and a fresh coating applied. Consequently that 


portion of the print which was thus retreated from time to time retained "new 
print quality" throughout the 35 bookings or approximately 164 runs. 

Remarks. It should be mentioned that the solvents employed are similar in 
inflammability to ethyl alcohol, and therefore all electrical equipment such as 
light fittings, motors, switches, etc., should be solvent-vapor-proof. Recircula- 
tion of the air in the cabinet is inadvisable. The exhaust vapors should be con- 
ducted outside the building. The solvents are similar to those used in quick- 
drying lacquers and finishes, and the same care should be exercised as when han- 
dling any inflammable volatile organic solvents. 

Acknowledgment. Acknowledgment is made to Mr. W. D. Kelly of Metro- 
Goldwyn-Mayer for the use of their prints for the preliminary field tests. Ac- 
knowledgment is likewise made to Mr. J. H. Spray of the Ace Film Laboratories, 
Brooklyn, for the practical information gained in his plant on the use of the lac- 
quer during the past year on color-films as well as on black-and-white negatives and 



QUESTION: What is the cost of applying the lacquer? 

* MR. TALBOT: The cost of materials is about $0.40 for each side for each 
1000-ft roll. The labor cost will vary from almost nothing, if the lacquer is ap- 
plied on the processing machine and no extra help is required to take care of it, up 
to $0.50 or $1.00 a roll if a special job is made of it and only a few rolls are treated. 

QUESTION: Can the lacquer be applied to various types of color-film? 

*MR. TALBOT: The difficulty with the application to color -film is that most color 
processes utilize dyes which are soluble either in water or dilute sodium carbonate 
which makes the removal of the lacquer difficult. It has been used commercially 
for some time by the Ace Laboratories for coating their duplitized color-prints 
largely for the purpose of the elimination of oil mottle, although the scratch resist- 
ance of the lacquer itself is also a factor. The new universal lacquer can be re- 
moved by an alcohol treatment, although care must be used to avoid warping the 
film. Experiments are under way with this new type of lacquer and the use of 
isopropyl alcohol for its removal from Technicolor film. 

The application and removal of the lacquer by carbonate solution is entirely 
satisfactory in the case of Kodachrome. 

QUESTION : What is the cost of equipment for applying the lacquer? 

*MR. TALBOT: That depends entirely upon the set-up. It is to be assumed 
that any processing laboratory will have equipment available, such as the stock 
roll, drying cabinet, and rewind mechanisms; therefore, the only special equip- 
ment is the coating unit. The cost of this coating unit will depend on the method 
employed. If the simple wick method is employed for speeds of 50 feet per min- 
ute or less, the cost of the equipment will be but a few cents, i. e., a strip of clean 
plush, a glass rod, and a lacquer pan. If the wick method is to be employed for 
speeds greater than 50 feet per minute, a plush-covered driven roll is necessary. 
The cost of this equipment will be the cost of a small motor plus the cost of a 
slotted roll for holding the plush. The cost of this roll should not exceed $5.00. 

For bead application the coating unit is somewhat more complicated and should 
be made with great precision, if the unit is to operate satisfactorily at high rates of 

* Communicated. 


speed. Detailed plans of such a unit are available from the Eastman Kodak Com- 
pany upon request. 

QUESTION: How many release prints could be made from a negative coated 
with this lacquer? 

*MR. TALBOT: The Ace Laboratories made 375 prints from a negative which 
had been coated on the emulsion surface with the protective film lacquer. It was 
not necessary to remove the lacquer from any reel during this printing. After the 
375 prints were made the coating was removed and the negative appeared to be in 
just as good condition as before release printing began. Presumably the nega- 
tive could be recoated and another 375 prints made. Possibly this cycle can be 
repeated many times. 

QUESTION: Is it necessary to use sodium carbonate solution for the removal 
of the lacquer or will any alkaline developer suffice? 

*MR. TALBOT: Laboratory tests indicate that any alkaline developer will re- 
move the lacquer in about two minutes and that the neutral rinse following the 
developer is not absolutely necessary. The use of a carbonate-formalin bath 
followed by a neutral rinse gives an additional factor of safety, but it is believed 
that the coating can be completely removed by passing the treated film through a 
commercial processing machine. 

QUESTION : Will the removal of the lacquer by developer harm the developer in 
any way? 

*MR. TALBOT: No, we believe not. The tests that we have run indicate that 
there is no change whatever in the action of the developer after its use to remove 
the lacquer. It would be necessary, however, to run this test on a much more 
extensive scale and with a wider variety of developers than we have used to be 
absolutely sure that no effect whatever occurs. 

QUESTION: Is the lacquer available in large quantities? 

*MR. TALBOT: The lacquer which the Ace Laboratories have been using for 
the emulsion side only is available in large quantities. The universal lacquer 
which is applicable to both sides of nitrate and safety films is available at the 
moment in sample lots only (one gallon), but unless something entirely unfore- 
seen happens, it will be available in large quantities in a few weeks. 



Since the introduction of the "Suprex" carbons about seven or eight years ago 1 
there has been a remarkable expansion in the use of these small-diameter, copper- 
coated carbons. The rapid growth of this type of arc is evidenced by its use in the 
majority of all the medium-size theaters in the country. The wide acceptance of 

* Presented at the 1940 Fall Meeting at Hollywood, Calif.; received October 
25, 1940. 

** National Carbon Company, Fostoria, Ohio. 

Feb., 1941] 



this high-intensity arc using a copper-coated non-rotating positive carbon and re- 
flector type lamp is largely due to the resultant brilliant snow-white light on the 
projection screen and to the economy of operation. Recently there have been 
introduced lamps and carbons which have extended the use of the high-intensity 
arc to the smallest theaters in the form of the "One-Kilowatt Arcs." 

Development work has continued in our laboratories on these non-rotating high- 
intensity carbons. This paper reports the progress on a new 8-mm copper-coated 
carbon of this type which, however, is not yet ready for distribution. This new 

70 Crush! ng 










70 Breaking 




30 2 














FIG. 1. Improvement in strength of carbon shells; 
experimental 8-mm shell over present 8-mm shell. 

carbon is stronger, produces light more efficiently, burns more steadily, and can 
be operated at higher currents than the present 8-mm "Suprex" positive. 


Strength of the Carbon. Carbon shells under stress deform only a small amount 
before they fracture, and therefore do not give the warning of the proximity of 
fracture that would be conveyed by a more deformable material of the same 
ultimate strength. This should be kept in mind in clamping the carbons in the 
carbon holders, which in most "Suprex" type lamps exert a powerful leverage 
capable of cracking the carbons if excess pressure is used. This is particularly 
true in the case of the positive carbon holder. The result of this cracking, which 



may be either transverse or longitudinal, is concealed by the copper coating but 
becomes evident when the carbon has been advanced in the holder and consumed 
to the point where the fracture exists. It is therefore of extreme importance that 
care be taken in clamping carbons in holders firmly but not excessively, and also 
in avoiding dropping carbons on the floor or otherwise mistreating them. 

This new experimental carbon has a shell with both a higher transverse break- 
ing strength and a higher crushing strength than the present one, as indicated in 
Fig. 1. Transverse breaking strength has been increased by 30 per cent, and the 
crushing strength, which is intimately connected with the action of the holder on 
the carbon, has been increased by 60 per cent. Therefore this increase in shell 
strength gives more assurance that these carbons will be free from breaks and 
cracks during the burning period. However, it does not mean that they are un- 


30 Life 




S6 65 


S6 G5 


S6 GS 

FIG. 2. Per cent improvement in operating characteristics of 
experimental 8-mm copper-coated high-intensity positive carbons 
over 8-mm "National" "Suprex." 

breakable, but that they have a substantially increased factor of safety with re- 
spect to breakage. 

Burning Performance. This experimental carbon has 24 per cent longer life at 
56 amperes, the lower limit of its current range, and 14 per cent longer life at 65 
amperes, which is the maximum current for which the present 8-mm "Suprex" 
carbon is rated. Moreover this is accomplished with only a 5-per cent reduction in 
screen light at 56 amperes and an actual increase of 8 per cent in screen light at 65 

If the amount of screen light is multiplied by the burning life, a measure of the 
efficiency of utilization of the carbon is obtained. For example, the product of the 
screen light in lumens and the burning life in hours per inch gives the total light 
energy in lumen-hours per inch of positive carbon. Fig. 2 shows that this experi- 
mental carbon gives materially more light energy per inch of positive carbon con- 
sumed, ranging from an 18-per cent increase at 56 amperes to a 23-per cent in- 
crease at 65 amperes. At 59 amperes the consumption of the experimental car- 
bon is the same as that of the present 8-mm "Suprex" at 56 amperes but the new 

Feb., 1941] 



carbon under these conditions gives 10 per cent more light. Similarly, at 63 am- 
peres the new carbon matches the consumption of the present "Suprex" at 60 
amperes but gives 10 per cent more light. These figures show that the user will 
be getting substantially more light energy per carbon. 

Arcs operated with the low- voltage power sources commonly used with "Su- 
prex" type lamps often show noticeable current fluctuation. This, as has been 
shown in an earlier publication, 2 is the result of a compensating action between the 
arc and the power source such that the current responds to momentary voltage 
changes in a manner tending to maintain a steady light output. The experi- 
mental carbon, particularly at the higher currents, gives a more stable arc and 
thereby eliminates to a large extent the necessity of compensating current fluctua- 
tions. Fig. 3 shows the superior stability at 65 amperes of the new carbon com- 
pared to the present 8-mm "Suprex." 

Qm/n 'Natlonal'Supraf 6mm &p. C.C HI. Postf/re 

FIG. 3. Records of the arc current showing the improved stability of the 
experimental 8-mm carbon over the 8-mm "National" "Suprex" positive. 

In some theaters now using "Suprex" carbons it would be desirable to obtain 
more light than can be obtained with the present 8-mm positive carbon at 65 am- 
peres. Higher currents have not been feasible because it has been found that the 
present 8-mm "Suprex" positive can not be operated much above its maximum 
current rating of 65 amperes without excessive current fluctuation. The experi- 
mental 8-mm positive carbon does not show this undesirable feature even at 70 
amperes and opens up the possibility of obtaining further increases in light. 

At 68 amperes the new carbon has the same consumption rate as the present 
"Suprex" at 65 amperes but gives about 20 per cent more light. At 70 amperes 
the new carbon delivers about 25 per cent more light with about 10 per cent higher 
consumption rate than the present "Suprex" at 65 amperes. When the current is 
increased from 65 to 70 amperes at the same arc length, the arc voltage increases 
about 2 volts. This increase in arc current and voltage might exceed the capacity 
of some of the power sources while the increased consumption is too great for most 
of the feed motors. However, the increase in voltage can be avoided by shortening 
the arc length about 0.05 inch, which will also reduce the consumption at 70 am- 


peres so that some of the lamps can feed the carbons rapidly enough. With these 
higher consumption rates, it is important that the negative be carried at its correct 
position because the crater face can become malformed very quickly if a poorly 
aligned negative is not corrected promptly. 

When low- voltage power sources designed for "Suprex"type lamps and carbons 
are used, the new carbon can be burned from 56 to 70 amperes subject at 70 am- 
peres to the limitations just described. In a few theaters there are still some old, 
high-voltage generators originally designed for Hi-Lo lamps. With these power 
sources, best results will be obtained with the new carbon if the current is main- 
tained at or above 60 amperes. 

This experimental 8-mm copper-coated high-intensity positive carbon has the 
best features of the present "Suprex" carbon and in addition has the advantages 
of greater strength, higher efficiency, steadier operation, and a wider current 
range, and therefore represents a significant advance over the present carbon. 


1 JOY, D. B., AND DOWNES, A. C.: "Direct- Current High-Intensity Arc with 
Non-Rotating Positive Carbons," J. Soc. Mot. Pict. Eng., XXII (Jan., 1934), No. 
1, p. 42. 

2 JOY, D. B., AND GEIB, E. R.: "The Non-Rotating High-Intensity D-C Arc 
for Projection," /. Soc. Mot. Pict. Eng., XXIV (Jan., 1935), No. 1, p. 47. 



The new screen, molded of plastic, represents a departure from the conven- 
tional. Its surface is not flat, but is smoothly contoured in a system of elliptical 
convexities forming a toric curve around each hole. Instead of being perforated 
in the usual manner, it is provided with holes molded in the shape of flaring horns, 
the sidewalls of the holes forming part of the surface contours. The screen has 
no seams, being molded as a single sheet in sizes up to 30 feet wide, with the plas- 
tic contoured to its three-dimensional pattern. 

This pattern is shown in Fig. 1, which represents the screen photographed nor- 
mal to the surface while illuminated by a flat light striking at a 10-degree inclina- 
tion. When Fig. 1 is observed for more than a few seconds, the phenomenon of 
alternate direct and reverse pseudoscopic vision will cause the contours to appear 
convex at one moment, concave the next. They are actually convex. One set of 
waves undulates on a right diagonal, the other on a left diagonal. A chevron 
pattern is formed by the position of the various crossing points of the two sets 
of waves. 

* Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received October 
20, 1940. 

** Walker- American Corp., St. Louis, Mo. 


Fig. 2 shows the structure. These molded contours are elliptical in cross-sec- 
tion, sweeping in a curve from the base of one hole to the base of the next adjacent 
hole, and forming a modified toric curve around each hole. 

Fig. 3 is a photomicrograph of a cross-section of the screen taken between two 
holes, showing the shape of the contour and sidewalls of the holes. The thick- 
ness is slightly under 0.025 inch, with the plastic surface bonded to a cloth back- 
ing approximately 0.010 inch thick. Forming the walls of each hole, the plastic 
extends in a thin ring through to the very back of the screen, thoroughly sealing 
the backing. 

While the single sheet of plastic that forms the surface is thick enough to sup- 
port its own weight, it requires reinforcement to prevent tearing under impact. 
Consequently, a strong cotton drill is impregnated with fire-retarding salts, a re- 

FIG. 1. Patterns of molded holes, photographed 
normal to the screen surface. 

sin, and a bonding agent. Placed under the plastic surface and thoroughly 
sealed, the structure is moisture-proof. This backing comprises strips 51 inches 
wide running vertically, and welded together at their edges in such a manner as 
to leave the frontal face of the backing smooth and flush under the one-piece plas- 
tic surface. 

The advantages of the plastic structure are many. Primarily, it is the only 
suitable medium in which these contours can be formed. Such a surface, molded 
as a single piece, possesses absolute uniformity from edge to edge of the entire 
screen. The customary seam, with its steadily increasing visibility, is eliminated. 
The durability of this hard, tough structure is obvious. There are no cut fibers to 
absorb dirt and moisture in and around the holes, thereby eliminating a great 
cause of deterioration. 

The smooth, hard surface, contoured in easy curves, provides no lodging places 
for dust and dirt. Tests show that this structure collects dirt at less than half the 


usual rate. Collection and irapaction of dirt in the holes, so detrimental to sound 
transmission, are virtually eliminated by the widely flaring horn-shaped holes, 
permitting maintenance of sound transmission at proper levels. 

The plastic employed is moderately flexible, hard, and tough at normal tem- 
peratures. It softens at temperatures above 140 F. When cold, at temperatures 
below 50 F, it becomes progressively brittle. It must not be unrolled or bent when 
cold, but once stretched, cold will not affect it adversely. This is the familiar 
characteristic of all thermoplastic materials loss of flexibility when cold. The 
screen is rolled for shipment in the usual manner, and is stretched by lacing within 
a frame in the customary fashion. Installation is no more difficult. While the 
plastic screen weighs three times as much as conventional types, and requires 
more care due to that weight, it is very easy to stretch flat, as it is molded without 

FIG. 2. Showing the structure of the screen. 
The shadows show the contour of the surface. 

wrinkles. Resistance to penetration is high. Damage from small missiles will be 
sharply reduced. Most small dents disappear in a few days. The screen is, of 
course, fire resistant. 

The open area at the bottom of the holes, through which light is lost, amounts to 
8.1 per cent of the total area. Overall reflective efficiency is 86 per cent, including 
the loss through the holes. The contour of the surface causes no sacrifice of light, 
as there is no baffle effect. It does improve distribution within the usable area. 
No disturbance of focus is caused, as the most critical projection lens has a depth 
of focus several hundred times the depth of this surface. 

The high reflective efficiency of the surface does not account for the visual effect 
produced. More of the light normally wasted in extreme side reflection is di- 
rected diffusely within the 120 degrees of the usable area. The plastic employed 
is transparent and nearly colorless, with a low refractive index. It is loaded by a 
new technic with a combination amorphous and crystalline pigment mixture pos- 


sessing a very high refractive index. The overall result is the properly balanced 
high spectral response necessary for correct color reproduction, with a wide lati- 
tude of tone and color. 

The shape of the holes, as shown by Fig. 3, is obviously beneficial to sound trans- 
mission. The air-column load and viscosity effect common to the normal perfora- 
tion is sharply reduced by the widely flaring sides of the molded hole. The rough 
ends of severed fibers and broken coating are replaced by a smooth plastic wall. 
The flaring horn shape of the hole permits improved angular distribution of sound. 

In theory, the elliptical contours of this surface should produce a different kind 
of picture. In practice, this has been the result. As one of these screens forms 
part of the projection installation at this meeting, opportunity has been pro- 
vided for observing what it delivers. The contoured imaging surface appears 
definitely to have influenced the illusion of depth. The "window effect" is in- 
creased. Technicolor shows the illusion at its best. Observers of a number of 
installations have commented upon a lack of flatness, apparent separation of fore 
and background to a greater degree, reduction in one's consciousness of the screen, 
and the effect of an image being formed in the air. While this surface plainly does 
not produce stereoscopic vision, a third dimension, or anything startling, it does 
produce an effect of depth that can not be evaluated but can be observed. 

FIG. 3. Photomicrograph of section of screen showing 
elliptical contour in profile. 

It is interesting to note the effect of this surface upon perspective, or the ability 
of the eye to estimate its position in space. If an observer will stand about two 
feet from a full-sized screen, far enough toward the center so that the masking is 
outside the line of vision, while the screen is illuminated by projected light or 
strong diffuse light, he will find difficulty in placing his finger within four inches of 
the correct plane of the screen after looking directly at it for thirty seconds. The 
finger will tend to select a point in space as the plane of the screen. 

Why should these tiny elliptical and toric contours influence a perspective illu- 
sion? Our theory has been based upon the psychology of monocular vision. 
Close one eye and one can still perceive solidity and perspective in the image 
formed on the retina. Helmholtz established the principle that recognition plays 
a part in the image perceived. The eye is accustomed to perspective, and it will 
build a mental image of perspective to aid the real image, if given the slightest 
assistance. This surface appears to provide something which aids that mental 

In this structure, 60 per cent of the picture falls upon the elliptically contoured 
area, and 40 per cent upon the toric curves surrounding each of the holes, thereby 
breaking the picture into a number of minute portions. These portions have 
varying angles in different image planes throughout the depth of the contours. 


As the entire structure is uniform, these portions upon reflection must be resolved 
into a series of total images lying in a multiplicity of planes. Certain portions of 
the contours cause a difference in the angular width of the image portions as pro- 
jected and as observed, due to the increase in the total surface area produced by 
the toric curves. The eye appears continuously to select certain minute char- 
acteristics found in normal perspective, integrate them with the image, and con- 
struct an added illusion of depth. 

This effect is an illusion of the picture as a whole, and it can be destroyed by 
distractions. Of course, it can not be observed upon a portion or sample. It can 
be found on a full screen only. Too much should not be expected, as it is only 
another contribution to beautiful projection, moving the goal of realism a step 


The camera referred to in the paper entitled "The Twentieth Century 
Camera and Accessories" by D. B. Clark and G. Laube, published in the 
January issue of the JOURNAL on p. 50 is now known as the "Cine-Simplex" 
camera and is being manufactured for general distribution by the Cine-Simplex 
Corp., Syracuse, N. Y. 



The editors present for convenient reference a list of articles dealing with subjects 
cognate to motion picture engineering published in a number of selected journals. 
Photostatic copies may be obtained from the Library of Congress, Washington, D. C., 
or from the New York Public Library, New York, N. Y. Micro copies of articles 
in magazines that are available may be obtained from the Bibliofilm Service, Depart- 
ment of Agriculture, Washington, D. C. 

American Cinematographer 

22 (January, 1941), No. 1 
Making Modern Night-Effects (pp. 11, 38-39) 
16-Mm Goes Professional (pp. 12, 34) 
South American Makes Movies (pp. 13-14, 39-40) 
Propose Standard Method of Determining Speed of 

Films (pp. 17, 40) 

Making Micro-Movies in 16-Mm Kodachrome (pp. 
24, 42^3) 

British Kinematograph Society, Journal 
3 (October, 1940), No. 4 

The 16-Mm Film as an Aid to Science (pp. 159-166) 
Recording 16-Mm Sound-Film from Double- Width 

Push-Pull Track (pp. 166-170) 
The Intimate Use of 16-Mm Films (pp. 170-175) 
Desirable Characteristics of Sub-Standard Projectors. 
(I) Heavy-Duty Models (pp. 176-179) 


20 (December, 1940), No. 12 

Extended Experimental Study of the Optical Pattern 
(pp. 22, 24) 

Educational Screen 

19 (December, 1940), No. 10 

Motion Pictures Not for Theaters, Pt. 22 (pp. 417- 

International Projectionist 

15 (November, 1940), No. 11 

Sound Heads on Parade: 1927-1940 (pp. 7-8, 11-12) 
The New "Orotip C" Negative Carbon for Low- 
Amperage Lamps (pp. 15-18) 












RCA's Fantasound System as Used for Disney's 

Fantasia (pp. 20-21, 24) 
New Walker Plastic Molded Screen Is Highly Efficient 

(pp. 26, 37-38) ' 

Kinematograph Weekly 

286 (December 5, 1940), No. 1755 
Projection with Sub-Standard Apparatus (pp. 9-10) 


13 (December, 1940), No. 12 

An Electrically Focused Multiplier Phototube (pp. 
20-23, 58, 60) 

International Photographer 

12 (January, 1941), No. 12 
Cooperative Research Laboratory Needed (pp. 8-9) 

Motion Picture Herald (Better Theaters Section) 
141 (December 14, 1940), No. 11 
Preparation for a New Advance in Motion Picture 
Sound (pp. 30, 32-33) 

Photographische Industrie 

38 (November 6, 1940), No. 45 

liber die Brauchbarkeit von Neophan-Filtern bei 
Farbenfilmaufnahmen (The Use of Neophan Filters 
in Color Photography) (pp. 664-666) 









Program and Facilities 

E. HUSE, President 

E. A. WILLIFORD, Past-President 

H. GRIFFIN, Executive Vice-President 

W. C. KUNZMANN, Convention Vice-President 

A. C. DOWNES, Editorial Vice-President 

G. A. BLAIR, Chairman, Local Arrangements 

S. HARRIS, Chairman, Papers Committee 

J. HABER, Chairman, Publicity Committee 

J. FRANK, JR., Chairman, Membership Committee 

H. F. HEIDEGGER, Chairman, Convention Projection 

Reception and Local Arrangements 





G. A. BLAIR, Chairman 
C. E. K. MEES 

Registration and Information 

W. C. KUNZMANN, Chairman 

Hotel and Transportation 

F. E. ALTMAN, Chairman 

Publicity Committee 

J. HABER, Chairman 






210 SPRING CONVENTION [j. s. M. P. E. 

Banquet and Dance 

I. L. NIXON, Chairman 





H. F. HEIDEGGER, Chairman 





Officers and Members Rochester Projectionists Local No. 253 


The headquarters of the Convention will be the Sagamore Hotel, where excellent 
accommodations and moderate rates are assured. A reception parlor will be pro- 
vided as headquarters for the Ladies' Committee. 

Hotel reservation cards will be mailed to the members of the Society early in 
April. They should be filled out and mailed immediately to the Sagamore Hotel 
so that suitable accommodations may be reserved, subject to cancellation if 
unable to attend the convention. 

The following European-plan day rates are extended by the Sagamore Hotel to 
Society members and guests attending the Convention (all rooms are outside 
rooms with bath) : 

Room for one person $3 . 00 to $5 . 00 

Room for two persons, double bed 4 . 50 to 6 . 00 

Room for two persons, twin beds 6.00 to 7.00 

Suite accommodations, one to four persons 12 . 00 and up 

The following hotel garage rates will be available to SMPE delegates and guests 
who motor to the Convention : 24-hr, inside parking, 75 ; outside parking (daily) , 

The colorful Sagamore Room on the main floor of the Hotel offers special break- 
fast, luncheon, and dinner menus at moderate prices. 

Golfing privileges at several Rochester country clubs may be arranged for 
either by the hotel management or at the SMPE registration headquarters. 


Convention registration and information headquarters will be located on the 
Sagamore Hotel roof, adjacent to the Glass House, where all technical sessions and 
symposiums will be held. 

Members and guests attending the Convention will be expected to register, and 
so help to defray the Convention expenses. Convention badges and identifica- 
tion cards will be provided for admittance to all regular and special sessions during 
the Convention. The identification card will also be honored through the 
courtesy of Loew's Theaters, Inc., at Loew's Rochester Theater and, through the 
courtesy of Monroe Amusements, Inc., at the Palace, Regent, and Century Theaters. 

Feb., 1941] SPRING CONVENTION 211 

Group visits to various plants in Rochester and vicinity may be arranged at 
the Registration Desk. 


All the technical sessions of the Convention will be held in the Glass House 
on the roof of the Sagamore Hotel with the exception of Wednesday evening, as 
described below. Members should note that the banquet, which at past con- 
ventions has always been held on Wednesday evening, this time has been sched- 
uled for Tuesday evening to permit holding a special meeting on Wednesday 
evening at the Eastman Theater. 

Wednesday, May 7th, will be devoted to a joint meeting of the Acoustical So- 
ciety of America and the SMPE, consisting of a symposium of papers by engineers 
of the Bell Telephone Laboratories in the morning and afternoon. In the evening 
a demonstration of stereophonic sound will be given by the Bell Telephone Labora- 
tories at the Eastman Theater. 


The usual Informal Get-Together Luncheon for members, their families, and 
guests will be held in the Starlight Room on the hotel roof on Monday, May 5th, 
at 12:30 P. M. : short addresses by prominent speakers; names to be announced 
later. Luncheon and banquet tickets should be procured when registering. 

The 48th Semi-Annual Banquet and Dance will be held in the Starlight Room 
on the hotel roof on Tuesday evening, May 6th, at 7:30 P. M.: music and enter- 
tainment. Banquet tickets should be procured and tables reserved at registra- 
tion headquarters by noon of Tuesday, May 6th. 


Mrs. C. M. Tuttle, Convention Hostess, and members of her Committee are 
arranging a very attractive program of entertainment for the ladies attending the 
Convention. A reception parlor will be provided for the use of the Committee 
during the Convention. 


Monday, May 5th 

9:00 a. m. Sagamore Hotel Roof 

9:30 a. m.-12:00 Glass House, Hotel Roof 

Technical session 
12:30 p. m. Starlight Room, Hotel Roof 

Get-Together Luncheon for members, their families, and 
guests. Brief addresses by several prominent speakers 
2:00 p. m. Glass House, Hotel Roof 

Technical session 
8:00 p. m. Glass House, Hotel Roof 

Technical session 



Tuesday, May 6th 
9:00 a. m. Sagamore Hotel Roof 

9:30 a. m. Glass House, Hotel Roof 

Society Business 

Technical session 
12:30 p. m. Luncheon period 

Program for this afternoon to be announced later 
7:30 p. m. Starlight Room, Hotel Roof 

Semi- Annual Banquet and Dance of the SMPE : addresses 

and entertainment: music, dancing, and entertainment 

Wednesday, May 7th 
10:00 a. m. Glass House, Hotel Roof 

Technical session 

12:30 p.m. Luncheon period 

2:00 p. m. Glass House, Hotel Roof 

Stereophonic sound papers session 
8:00 p. m. Eastman Theater 

Stereophonic sound demonstration for the SMPE Conven- 
tion and invited groups. Admission only by SMPE 
identification card, or special invitation card 

Thursday, May 8th 
10:00 a. m. Glass House, Hotel Roof 

Technical session 

12:30 p. m. Luncheon period 

2:00 p. m. Glass House, Hotel Roof 

Technical and business session 



Convention Vice- President 



At a meeting of the Section held at the Hotel Pennsylvania, New York, N. Y., 
on January 15th, Mr. Lawrence R. Martin, Production Technician, Camera 
Works, Eastman Kodak Company, Rochester, N. Y., presented a paper on 
"The Motion Picture as a Tool in Science and Engineering." 

Mr. Martin discussed applications of the motion picture apart from the usual 
role the motion picture plays in the theater as a vehicle of entertainment. Methods 
were described of recording the indications of aeroplane panel instruments, 
reactions of the aeroplane operator while flying, irregularities of motion in pro- 
duction machinery, calibration of special gasoline storage tanks, and study of 
the action of locomotive wheels upon the rails. Sixteen-mm pictures showing 
these various applications were projected. 

The next meeting of the Section will be held on February 19th. 


As the result of balloting by the Standards Council of the American Standards 
Association, the Standards and Recommended Practices listed below have been 
approved by the A. S. A. as of January 10, 1941. It is expected that these Stand- 
ards and Recommended Practices will be published in the March issue of the 

American Standards 

35-mm Sound Film; Emulsion and Sound Record Positions in Camera 

35-mm Sound Film; Emulsion and Sound Record Positions in Projector 

Positive (for Direct Front Projection). 
35-mm Film; Projection Reels. 
16-mm Silent Film; Cutting and Perforating Negative and Positive Raw 


16-mm Film; Projector Sprockets. 
16-mm Silent Film; Camera Aperture. 
16-mm Silent Film ; Projector Aperture. 

16-mm Silent Film; Emulsion Position in Camera Negative. 
16-mm Silent Film; Emulsion Position hi Projector Positive (for Direct 

Front Projection). 
16-mm Film; Projection Reels. 
16-mm Silent Film; Cutting and Perforating Negative and Positive Raw 


16-mm Sound Film; Camera Aperture. 
16-mm Sound Film; Projector Aperture. 



16-mm Sound Film; Emulsion and Sound Record Positions in Camera 

16-mm Sound Film; Emulsion and Sound Record Positions in Projector 


8-mm Film, Cutting and Perforating Negative and Positive Raw Stock. 
8-mm Film; 8-Tooth Projector Sprockets. 
8-mm Silent Film; Camera Aperture. 
8-mm Silent Film; Projector Aperture. 

8-mm Silent Film; Emulsion Position in Camera Negative. 
8-mm Silent Film; Emulsion Position in Projector Positive (for Direct 

Front Projection). 
8-mm Silent Film; Projection Reels. 

American Recommended Practices 

16-mm Silent Film; Film Splices Negative and Positive. 

16-mm Sound Film; Film Splices Negative and Positive. 


Photographic Density. 

Projection Rooms. 

Projection Screens. 


Safety Film. 

Fader Setting Instructions. 

Nomenclature for Filters. 


The Tenth Annual Meeting of the Council will be held March 3 to 5, 
1941, at Washington, D. C. The technical session will be co-sponsored by the 
American Society for Testing Materials, and will be held on March 5th during 
their eastern regional meetings. Delegates and members of the Council are 
urged to be present, and members of all member-bodies and any others interested 
in color are invited to attend the sessions. Information concerning the detailed 
program may be obtained by request from the Secretary of the I. S. C. C. (Box 
155, Benjamin Franklin Station, Washington, D. C.) after February 1st. 




Volume XXXVI March, 1941 



American Motion Picture Standards and Recommended Prac- 
tices 217 

Report of the SMPE Standards Committee 260 

Theater Acoustic Recommendations of the Academy Research 
Council Theater Standardization Committee 267 

An Outline of the Work of the Academy Research Council Sub- 
Committee on Acoustical Characteristics J. DURST 283 

Twenty- Four Years of Service in the Cause of Better Projection 


Line Microphones H. F. OLSON 302 

New Motion Picture Apparatus 

A Line Type of Microphone for Speech Pick-Up. 


Current Literature 315 

1941 Spring Convention at Rochester, N. Y., May 5th to 8th, 
Inclusive 317 

Society Announcements 321 





A. C. DOWNES, Chairman 




Subscription to non-members, $8.00 per annum; to members, $5.00 per annum, 
included in their annual membership dues; single copies, $1.00. A discount 
on subscription or single copies of 15 per cent is allowed to accredited agencies. 
Order from the Society of Motion Picture Engineers, Inc., 20th and Northampton 
Sts., Easton, Pa., or Hotel Pennsylvania, New York, N. Y. 

Published monthly at Easton, Pa., by the Society of Motion Picture Engineers. 

Publication Office, 20th & Northampton Sts., Easton, Pa. 

General and Editorial Office, Hotel Pennsylvania, New York, N. Y. 

West Coast Office, Suite 226, Equitable Bldg., Hollywood, Calif. 

Entered as second class matter January 15, 1930, at the Post Office at Easton, 
Pa., under the Act of March 3, 1879. Copyrighted, 1941, by the Society of 
Motion Picture Engineers, Inc. 


**President: EMERY HUSE, 6706 Santa Monica Blvd., Hollywood, Calif. 
**Past-President: E. ALLAN WILLIFORD, 30 East 42nd St., New York, N. Y. 
** Executive Vice-President: HERBERT GRIFFIN, 90 Gold St., New York, N. Y. 

*Engineering Vice-President: D. E. HYNDMAN, 350 Madison Ave., New York. 
** Editorial Vice-President: ARTHUR C. DOWNES, Box 6087, Cleveland, Ohio. 

^Financial Vice-President: A. S. DICKINSON, 28 W. 44th St., New York, N. Y. 
** Convention Vice-President: W. C. KUNZMANN, Box 6087, Cleveland, Ohio. 

* Secretary: PAUL J. LARSEN, 44 Beverly Rd., Summit, N. J. 

* Treasurer: GEORGE FRIEDL, JR., 90 Gold St., New York, N. Y. 


**M. C. BATSEL, Front and Market Sts., Camden, N. J. 

*J. A. DUBRAY, 1801 Larchment Ave., Chicago, 111. 

*JOHN G. FRAYNE, 6601 Romaine St., Hollywood, Calif. 

*A. N. GOLDSMITH, 580 Fifth Ave., New York, N. Y. 

*ARTHUR C. HARDY, Massachusetts Institute of Technology, Cambridge, Mass. 
**L. L. RYDER, 5451 Marathon St., Hollywood, Calif. 

*T. E. SHEA, 195 Broadway, New York, N. Y. 

*R. O. STROCK, 35-11 35th St., Astoria, L. I., N. Y. 

Term expires December 31, 1941. 
**Term expires December 31, 1942. 



The following American Standards and Recommended Practices 
(Z22.2 to Z22.35) comprise the entire body of motion picture proj- 
ects approved by the American Standards Association, and in force 
at the present date. Z22.34 and Z22.35 were approved Sep- 
tember 20, 1930; all the others listed below were approved 
January 10, 1941, either as revisions of previous specifications or as 
entirely new standards or recommended practices. In some in- 
stances the revisions consist only in changes of Z-numbers. The 
present publication supersedes all previous publications of American 
Motion Picture Standards and Recommended Practices. 


Z22 . 2 35-Mm Sound Film ; Emulsion and Sound Record Position in Camera 

Z22.3 35-Mm Sound Film; Emulsion and Sound Record Positions in Pro- 
jector Positive. For Direct Front Projection. 

Z22.4 35-Mm Film; Projection Reels 

Z22.5 16-Mm Silent Film; Cutting and Perforating Negative and Positive 
Raw Stock 

Z22.6 16-Mm Film; Projector Sprockets 

Z22.7 16-Mm Silent Film; Camera Aperture 

Z22.8 16-Mm Silent Film; Projector Aperture 

Z22.9 16-Mm Silent Film; Emulsion Position in Camera Negative 

Z22.10 16-Mm Silent Film; Emulsion Position in Projector Positive. For 
Direct Front Projection 

Z22.ll 16-Mm Film; Projection Reels 

Z22.12 16-Mm Sound Film; Cutting and Perforating Negative and Positive 
Raw Stock 

Z22.13 16-Mm Sound Film; Camera Aperture 

Z22. 14 16-Mm Sound Film; Projector Aperture 

Z22 . 15 16-Mm Sound Film ; Emulsion and Sound Record Positions in Camera 

Z22.16 16-Mm Sound Film; Emulsion and Sound Record Positions in Pro- 
jector Positive 

Z22.17 8-Mm Film; Cutting and Perforating Negative and Positive Raw 



Z22. 18 8-Mm Film; 8-Tooth Projector Sprockets 

Z22.19 8-Mm Silent Film; Camera Aperture 

Z22.20 8-Mm Silent Film; Projector Aperture 

Z22.21 8-Mm Silent Film; Emulsion Position in Camera Negative 

Z22.22 8-Mm Silent Film; Emulsion Position in Projector Positive. For 

Direct Front Projection 

Z22.23 8-Mm Silent Film; Projection Reels 
Z22.34 35-Mm Film; Cutting and Perforating Negative and Positive Raw 

Z22.35 35-Mm Film; 16-Tooth Projector Sprockets 


Z22.24 16-Mm Silent Film; Film Splices Negative and Positive 

Z22.25 16-Mm Sound Film; Film Splices Negative and Positive 

Z22 . 26 Sensitometry 

Z22.27 Photographic Density 

Z22 . 28 Projection Rooms 

Z22.29 Projection Screens 

Z22 . 30 Nomenclature 

Z22.31 Safety Film 

Z22 . 32 Fader Setting Instructions 

Z22.33 Nomenclature for Filters 

Unit of Photographic Intensity 

The formulation of these Standards and Recommended Practices 
is the result of effective collaboration over a long period by a large 
number of important groups in the motion picture industry. The 
continued and valuable cooperation of the Research Council of the 
Academy of Motion Picture Arts and Sciences, and of the Society of 
Motion Picture Engineers are gratefully acknowledged. The work 
of preparation and the method of approval were in accordance with 
the Sectional Committee procedure of the American Standards 
Association. It will be noted that no references are given in the 
standards or recommended practices to the originating sources of 
such material. Where specific publications are mentioned as a 
matter of information, the references are given in footnotes where the 
complete publications in question could not conveniently be included 
in the text. 

Mar., 1941] 



The membership of the Sectional Committee on Motion Pictures 
(Z22) is as follows: 

Academy of Motion Picture Arts and Sciences 

Acoustical Society of America 
Agfa Ansco Corporation 
Akeley Camera Company 
Amateur Cinema League, Inc. 

American National Committee for International Con- 
gresses of Photography 
American Society of Cinematographers 

Bell & Howell Company 

Dupont Film Manufacturing Corporation 

Eastman Kodak Company 

Electrical Research Products, Inc. 
Fire Protection Group of the ASA 

Illuminating Engineering Society 

International Projector Corporation 

Mitchell Camera Company 

Motion Picture Producers and Distributors of America 

National Carbon Company 

National Electrical Manufacturers Ass'n 
Optical Society of America 
RCA Manufacturing Company 
Society of Motion Picture Engineers 

Theater Equipment Supply Manufacturers Ass'n 
U. S. Bureau of Standards 



* (R. G. HOLSLAG) 







* (G. W. BOOTH) 




D. B. Joy 



O. F. NEU 

E. W. ELY 

January 12, 1941 




U. S. M. P. E. 

For 35 mm Sound Film 





m a^ ^eew from inside the camera looking 
toward the camera lens. 

(1) Emulsion position in camera: toward the lens, except for 

special processes. 

(2) Speed: 24 frames per second. 

(3) Distance between center of picture and corresponding sound: 

20 frames. 

Mar., 1941] 



For 35 mm Sound Film 



For Direct Front Projection 




Drawing shows film as seen from the light-source in the projector. 

(1) Emulsion position in projector: toward the light-source, except 

for special processes. 

(2) Speed : 24 frames per second. 

(5) Distance between center of picture and corresponding sound : 
20 frames. 



U. S. M. P. E. 

For 35 mm Film 




300 Meters 

1000 Feet 

600 Meters 

2000 Feet 








8.3 Min. 

0.328 Min. 

Recommended Practice 







* This dimension applies only within a radius of 0.5 inch from the axis of the 

Mar., 1941] 



For 16 mm Silent Film 







Inch Equivalents 


16.00 +0.00 

0.630 +0.000 

- 0.05 

- 0.002 


7.620 0.013 

0.3000 0.0005 


1.83 0.01 

0.0720 0.0004 


1.27 0.01 

0.0500 0.0004 


1.83 0.05 

0.072 0.002 


12.320 0.025 

0.485 0.001 


Not > 0.025 

Not > 0.001 


762.00 0.76 

30.0 0.03 




* L = the length of any 100 consecutive perforation intervals. 

These dimensions and tolerances apply to the material immediately 
after cutting and perforating. 



LT. S. M. P. E. 

For 16 mm Film 



I ! 

Number of Teeth in Mesh 
















600' * 0.5' 
450' =*= 0.5' 
300' 0.5' 
2230' * 0.5' 


























600' * 0.5' 
450' * 0.5' 
300' 0.5' 
2230' =*= 0.5' 











600' * 0.5' 
450' * 0.5' 
30 8 0' * 0.5' 
2230' * 0.5' 









For All 
1 Sprockets 





Inch Equivalents 

N = Number of teeth on sprocket. 
Tolerance for B and C +0 . 000 to - . 025 mm. 
or +0.000 to -0.001 in. 
Dimensional standards indicated by capital 
Recommended practice indicated by lower 
case letters. 
Values of C are omitted in cases where the 
angle of wrap on the sprocket would exceed 

12.22 + 0.05 
- 0.00 
1.22 + 0.00 
- 0.08 
B-0.3, Max. 
B + 1.52, 

0.481 + 0.002 
- 0.000 
0.048 + 0.000 
- 0.003 
B-0.01, Max. 
B + 0.060, 

Mar., 1941] 



For 16 mm Silent Film 






Inch Equivalents 


10.41 0.05 
7.47 0.05 
8.00 * 0.05 
0.5 approx. 

0.410 0.002 
0.294 0.002 
0.315 =*= 0.002 
. 02 approx. 

a = b = V longitudinal perforation pitch. 

These dimensions and locations are shown relative to unshrunk 
raw stock. 

The center of the frame-line shall pass through the centers of per- 
forations on opposite sides of the film. 



LT. S. M. P. E. 

For 16 mm Silent Film 






Inch Equivalents 



9.65 0.05 
7.21 =t 0.05 
8.00 * 0.05 
0.5 approx. 

0.380 * 0.002 
0.284 =fe 0.002 
0.315 0.002 
0.02 approx. 

a = b = V longitudinal perforation pitch. 

These dimensions and locations are shown relative to unshrunk 
raw stock. 

The center of the frame-line shall pass through the centers of 
perforations on opposite sides of the film. 

Mar., 1941] 



For 16 mm Silent Film 




Drawing shows film as seen from inside the camera looking 
toward the camera lens. 

(1) Emulsion position in camera: toward the lens, except for 

special processes. 

(2) Normal speed: 16 frames per second. 



tf. S. M. P. E. 

* For 16 mm Silent Film 

For Direct Front Projection 




Drawing shows film as seen from the light-source in the projector. 

(1) Emulsion position in projector: toward the lens, except for 

special processes. 

(2) Normal speed: 16 frames per second. 

Mar., 1941] 



For 16 mm Film 




120 Meters 

400 Feet 

240 Meters 

800 Feet 

480 Feet 

1600 Feet 


Inch Equivalents 


Inch Equivalents 


Inch Equivalents 

8.10 + 0.00 
- 0.08 
8.10 + 0.00 
- 0.08 
17.2 Min. 

0.319 + 0.000 
- 0.003 
0.319 + 0.000 
- 0.003 
0.677 Min. 

8.10 + 0.00 
- 0.08 
8.10 + 0.00 
- 0.08 
17.2 Min. 

0.319 + 0.000 
- 0.003 
0.319 + 0.000 
- 0.003 
0.677 Min. 

8.10 + 0.00 
8.10 + 0.00 
- 0.08 
17.2 Min. 

0.319 + 0.000 
- 0.003 
0.319 + 0.000 
- 0.003 
0.677 Min. 

Recommended Practice 








NOTE: Center Spindle Holes Either a combination of square and round holes or two square 
holes may be used. 



[j. S. M. P. E. 

For 16 mm Sound Film 




-H E 



Inch Equivalents 



16.00 +0.00 
- 0.05 
7.620 0.013 
1.83 0.01 
1.27 0.01 
1.83 0.05 
762.00 0.76 

0.630 +0.000 
- 0.002 
0.3000 0.0005 
0.0720 0.0004 
0.0500 0.0004 
0.072 0.002 
30.00 0.03 

*L = the length of any 100 consecutive perforation intervals. 

These dimensions and tolerances apply to the material immediately 
after cutting and perforating. 

Mar., 1941] 



For 16 mm Sound Film 












or FILM 


Inch Equivalents 



10.41 =*= 0.05 
7.47 0.05 
8.00 0.05 
0.5 approx. 

0.410 =*= 0.002 

0.294 0.002 
0.315 * 0.002 
. 02 approx. 

a = b = l /t longitudinal perforation pitch. 

These dimensions and locations are shown relative to unshrunk 
raw stock. 

The center of the frame-line shall pass through the center of a 



U. S. M. P. E. 


For 16 mm Sound Film 







Inch Equivalents 


9.65 0.05 
7.21 =*= 0.05 
8.00 =*= 0.05 
. 5 approx. 

0.380 =*= 0.002 
0.284 0.002 
0.315 =*= 0.002 
. 02 approx. 

a = b = Va longitudinal perforation pitch. 

These dimensions and locations are shown relative to unshrunk 
raw stock. 

The center of the frame-line shall pass through the center of a 

Mar., 1941] 



For 16 mm Sound Film 




Drawing shows film as seen from inside the camera looking toward the 
camera lens. 

(1) Emulsion position in camera: toward the lens, except for 

special processes. 

(2) Speed: 24 frames per second. 

(3) Distance between center of picture and corresponding sound: 

26 frames. 



LT. s. M.'P.JE. 

For 16 mm Sound Film 

For Direct Front Projection 




Drawing shows film as seen from the light-source in the projector. 

(1) Emulsion position in projector: toward the lens, except for 

special processes. 

(2) Speed : 24 frames per second. 

(3) Distance between center of picture and corresponding sound: 

26 frames. 

Mar., 1941] 



For 8 mm Film 










16.00 + 0.00 

0.630 + 0.000 

- 0.05 

- 0.002 


3.810 0.013 

0.150 * 0.0005 


1.83 * 0.01 

0.072 0.0004 


1.27 0.01 

0.0500 * 0.0004 


1.83 * 0.05 

0.072 0.002 


12.320 0.025 

0.485 0.001 


Not > 0.025 

Not > 0.001 


8.00 + 0.00 

0.315 + 0.000 

- 0.08 

- 0.003 


381.00 0.38 

15.000 =t 0.015 




* L = the length of any IOC consecutive perforation intervals. 

These dimensions and tolerances apply to the material immediately after cut- 
ting and perforating. 

Film may be slit before or after processing. 



[J. S. M. P. E. 

For 8 mm Film 









5.72 =*= 0.03 
9.42 + 0.00 
- 0.05 
1.02 + 0.00 
- 0.05 
1.14 + 0.08 
- 0.00 

450' = 

0.225 0.001 
0.371 + 0.000 
- 0.002 
0.040 + 0.000 
- 0.002 
0.045 + 0.003 
- 0.000 

t 0.5' 

Recommended Practice 





Mar., 1941] 



For 8 mm Silent Film 




<fe OF FILM 



Inch Equivalents 


4.80 * 0.03 
3.51 =*= 0.03 
5.22 =*= 0.05 

0.189 0.001 

0.138 =*= 0.001 
0.205 0.002 

a = b = l /2 longitudinal perforation pitch. 



tf. S. M. P E. 



For 8 mm Silent Film 









4.37 =*= 0.03 
3.28 =*= 0.03 
5.22 0.05 

0.172 =>= 0.001 
0.129 == 0.001 
0.2055 =*= 0.002 

a = b = l / 2 longitudinal perforation pitch. 

Mar., 1941] 





For 8 mm Silent Film 





Drawing shows film from inside the camera, looking toward the camera lens. 

(1) Emulsion position in camera: toward the lens, except for special 

(2) Normal speed: 16 frames per second. 



D". S. M. p. E. 

For 8 mm Silent Film 

For Direct Front Projection 









Drawing shows film as seen from the light-source in tl*e projector. 

(1) Emulsion position in projector: toward the lens, except for 
special processes. 

(2) Normal speed: 16 frames per second. 

Mar., 1941] 



For 8 mm SUent Film 



Capacity, 60 M. (200 Ft.) 


Inch Equivalents 


8.10 + 0.00 
- 0.08 

0.319 + 0.000 
- 0.003 

Recommended Practice 





Drive side of sprocket may have any desired odd number of driv- 
ing slots, evenly spaced. 



LT. S. M. P. E. 

For 16 mm Silent Film 








Inch Equiv. 


Inch Equiv. 






Mar., 1941] 



For 16 mm Sound Film 







Inch Equiv. 


Inch Equiv. 








[J. S. M. P. E. 






The principle of non-intermittency shall be adopted as recom- 
mended practice in making sensitometric measurements. 







The integrating sphere shall be used as a primary instrument for 
the determination of photographic density. Photographic densities 
determined by means of this primary instrument shall be used as 
secondary or reference standards by means of which densitometers 
of other types may be calibrated. 







Projection Lens Height. The standard height from the floor to 
the center of the projection lens of a motion picture projector should 
be 48 inches. 

Projection Angle. Should not exceed 12 degrees. 

Observation Port. Should be 12 inches wide and 14 inches high, 
and the distance from the floor to the bottom of the openings shall 
be 48 inches. The bottom of the opening should be splayed 15 de- 
grees downward. If the thickness of the projection room wall 
should exceed 12 inches, each side should be splayed 15 degrees. 

Projection Lens Mounting. The projection lens should be so 
mounted that the light from all parts of the aperture shall traverse 
an uninterrupted part of the entire surface of the lens. 

Projection Lens Focal Length. The focal length of motion picture 
projection lenses should increase in Y-rinch steps up to 8 inches, 
and in x /2-inch steps from 8 to 9 inches. 

Projection Objectives, Focal Markings. Projection objectives 
should have the equivalent focal length marked thereon in inches, 
quarters, and halves of an inch, or in decimals, with a plus (+) or 
minus ( ) tolerance not to exceed 1 per cent of the designated focal 
length also marked by proper sign following the figure. 

(Complete plans for projection rooms are contained in /. Soc. Mot. Pict. Eng. t 
Nov., 1938, p. 484.) 

Mar., 1941] 








Sizes of screens shall be in accordance with the table below. 

The spacing of grommets shall be 6 inches, with 12 inches as a 
possible sub-standard. The ratio of width to height of screens shall 
be 4 to 3. 

The width of the screen should be equal to approximately 1 / 6 
the distance from the screen to the rear seats of the auditorium. 
The distance between the front row of seats and the screen should 
be not less than 0.87 foot for each foot of screen width. 

Screen Sizes 

Size No. 
of Screen 




Size No. 
of Screen 

































































































































LT. S. M. P. E. 






Number of Teeth in Mesh. The number of teeth in mesh with the 
film (commonly referred to as "teeth in contact") shall be the number 
of teeth in the arc of contact of the film with the drum of the sprocket 
when the pulling face of one tooth is at one end of the arc. 







Safety Film. The term "Safety Film," as applied to motion pic- 
ture materials, shall refer to materials having a burning time greater 
than 10 seconds and falling into the following classes: (a) support 
coated with emulsion, (b) any other material upon which or in which 
an image can be produced, (c) the processed products of these ma- 
terials, and (d) uncoated support that is or can be used for motion 
picture purposes in conjunction with the aforementioned classes of 

The burning time is denned as the time in seconds required for the 
complete combustion of a sample of the material 36 inches long, the 
determination being according to the procedure of the Underwriters 
Laboratory. This definition was designed specifically to define Safety 
Film in terms of the burning rate of the commercial product of any 
thickness or width used in practice. The test of burning time, there- 
fore, shall be made with a sample of the material in question having 
a thickness and width at which the particular material is used in 

All 16 and 8-mm film must be of the safety type. 







The Fader Setting Instruction Leader shall consist of 15 frames 
located in the first 20 frames of the synchronizing leader; the first 
frame shall designate the type of print; the second frame the type 
of reproducing equipment necessary to project the print; and the 
next nine frames the general fader setting specified in relation to an 
average fader setting for the particular product under consideration. 
The remaining frames may be used for whatever additional informa- 
tion the studio may wish to transmit to the theater. 

The designation "Regular" in the Instruction Leader indicates 
that only one type of print has been issued on the particular produc- 
tion under consideration. Productions with prints designated as 
either "Hi-Range" or "Lo-Range" are issued in both types of prints, 
i. e., all productions on "Hi-Range" prints will have necessarily been 
issued on "Lo-Range" prints as well. 

Both the terms "push-pull" and "single" shall be on every leader, 
one or the other being crossed out to leave the proper term desig- 
nating the type of sound-track on the print. 

Mar., 1941] 











The symbol describing any filter shall consist of three characters, 
the first designating the frequency of 3-db insertion loss, the second 
the character "Hi" or "Lo" to indicate high-pass or low-pass, and 
the third the frequency of 10-db insertion loss (all frequencies in 

Thus the following describes several low-pass filters "4000 Lo 
6000" (Fig. 1), "5000 Lo 7000" or "4500 Lo 5500" and the following 
describe several high-pass filters: "60 Hi 40" (Fig. 2), "80 Hi 30," 
or "100 Hi 50." 

A combination of two of the above symbols may be used to de- 
scribe a band-pass filter (Fig. 3) or a dividing network (Fig. 4) or a 
reverse combination of symbols may be used to describe a band- 
elimination filter (Fig. 5). 

Mar., 1941] 



1 III! 







ricuMC t- LOW-PA: 












7 -Ml -40 













)WO K 





> 1000 JC 




: 3 -BAND-PASS rn.TCR 








^ - 



^ . 





















rieunc >-aAND-aji 







[J. S. M. P. E. 

For 35 mm Film 


















J_ rfa 



T ^r 

G ._d-j_ 






D R 


rv . 







7ncA Equivalents 


35.00 + 0.00 

1.378 +0.000 

- 0.05 

- 0.002 


4.750 0.013 

0.1870 == 0.0005 


2.794 0.01 

0.1100 0.0004 


1.98 0.01 

0.0780 0.0004 


3.40 0.05 

0.134 0.002 


28.17 == 0.05 

1.109 0.002 


Not > 0.025 

Not > 0.001 


475.00 0.38 

18.700 0.015 




* L = the length of any 100 consecutive perforation intervals. 

These dimensions and tolerances apply to the material immediately 
after cutting and perforating. 

Mar., 1941] 



For 35 mm Film 



Feed Sprocket 

Intermittent Sprocket 

Take-Up (Hold- Back) Sprocket 





. Millimeters 






27.36 0.03 
24 . 00 o . 03 
1.40 + 0.00 
- 0.05 
1 . 40 . 00 
- 0.05 

1.097 0.001 
0.945 0.001 
0.055.+ 0.000 
- . 002 
0.055 + 0.000 
- . 002 

27.86 0.03 
24 . 00 . 005 
1.40 + 0.00 
- 0.05 
1.40 + 0.00 
- 0.05 

1.097 0.001 
0.945 0.0002 
0.055 + 0.000 
- 0.002 
0.055 + 0.000 
- . 002 

27.86 0.03 
23.67 0.03 
1.40 + 0.00 
- 0.05 
1 . 40 + . 00 
- 0.05 

1.097 0.001 
0.932 0.001 
. 055 + . 000 
- . 002 
0.055 + 0.000 
- 0.002 

22 Deg. 30 Min. 1.5 Min. 

22 Deg. 30 Min. 0.75 Min.* 

22 Deg. 30 Min. 1.5 Min. 

Recommended Practice 










* The accumulated error between any two teeth not to exceed 4 minutes. 




While no standard or recommended practice for a unit of photo- 
graphic intensity is proposed at this time, the description of an 
international unit of photography as adopted by the International 
Congress of Photography in 1928 and 1931 is here given as a matter 
of information. The resolution passed by the 1928 Congress defining 
the unit of photographic intensity received the approval of the 
English and American national committees and of the Optical 
Society of America. At the 1931 Congress an amendment to the 
resolution passed by the 1928 Congress was proposed and accepted 
by the representatives of the various national committees in at- 
tendance. So far as can be determined, no official approval by the 
various national committees was subsequently given. There is 
therefore a little doubt as to the exact status of this standard as 
established by the International Congress. The ASA Sectional 
Committee Z22, Motion Pictures, is at the present time in active 
cooperation with the ASA Sectional Committee Z38, Photography, 
in an endeavor to clarify this matter and to formulate a proposal 
for an American standard for a unit of photographic intensity. 

The unit of photographic intensity for the sensitometry of nega- 
tive materials may be defined as the intensity of a filtered source of 
radiation having a luminous intensity of one international candle, 
and produced by a gray body at a color temperature of 2380 (accord- 
ing to the most recent determination of the international tempera- 
ture scale), together with a selectively absorbing filter made up as 
follows: Two solutions compounded according to the following 
formula, the complete filter to consist of a 1-cm* layer of each solu- 
tion contained in a double cell made by using three pieces of boro- 
silicate crown glass (refractive index, D = 1.51), each 2.5 mm thick. 

* Tolerance in thickness shall be 0.05 mm. 


Solution A 

Copper sulfate, (CuSO 4 , 5H 2 O) 3.707 gm* 

Mannite, (C 6 H 8 (OH) 6 ) 3 . 707 gm* 

Pyridine, (C 5 H 5 N) 30.0 cc 

Water (distilled) to make 1000. cc . 

Solution B 
Cobalt ammonium sulphate, 

(CoSO 4 , (NH 4 ) 2 S0 4 , 6H 2 0) 26.827 gm* 

Copper sulfate, (CuSO 4 , 5H 2 O) 27. 180 gm* 

Sulfuric acid (sp. gr. 1.835) 10.0 cc 

Water (distilled) to make 1000. cc 

The spectrophotometric absorption characteristics of the filter 
made up according to these specifications are shown in the following 
chart,** and in the table below this information is given in numerical 

The unit of photographic intensity as recommended by the Seventh Inter- 
national Congress of Photography, held in London in July, 1928, was formally 
approved by the English and American Committees of the Congress and by the 
Optical Society of America. In ratifying this standard, the Optical Society of 
America, in order to forestall any possible misinterpretation of the intent of the 
resolution, presented the following clarifying statements: 

"In ratifying this proposal the Optical Society of American understands that the 
intent of this recommendation is as follows: 

(V) The intention is to specify two things, (a} the unit in which the intensities 
of light-sources are to be expressed, and (b) the quality of light to be used. 

(2} The unit is to be the International Candle, implying further that the in- 
tensities measured and stated will be luminous intensities as in visual photometry. 

(5) The quality of light to be used for sensitometry of negative materials is to 
be that which results from passing the radiation from a gray body at 2380 K nor- 
mally through the filter described. 

(4) The gray body and the selectively absorbing filter together shall be con- 
sidered as the effective source in specifying the intensity (candle-power). 

* For practical purposes an accuracy to the second place of decimals is prob- 
ably sufficient. 

** Bur. Standards Misc. Pull. No. 114: Raymond Davis and K. S. Gibson, 
"Filters for the Reproduction of Sunlight and Daylight, and the Determination 
of Color-Temperature." 



[J. S. M. P. E. 

ISOj i i [ i i i 
125 H 


~ - ^ 

_ ^ 





_ i j. 

25 Z - 2 2 r - 
( r ~ ~ ^ 

-. i 

~ L 

- ? - 

^ V* 

- r: 



= s; 

;- - = 




550 400 450 500 550 60O 650 700 7' 

Wavelength millimicrons (m/x) 


































































































































































T Spectral Transmission of Filter at 

25 C 

V Relative Visibility Function 
E Relative Energy of 2360 K 
E' Relative Energy of Mean Noon 

Sunlight at Washington 

E"(= T X E)* Relative Energy of 
2380 K and Filter Combination 

Light Transmission of Filter 
for 2380 K = 0.1352** 

* Adjusted to make sum of E" E' from 400 to 720 m? equal practically to 


** Factor to be used to multiply the candle-power of the light-source to obtain 
the candle-power of the source-and-filter combination. 


(5) The procedure recommended for determining the intensity of the com- 
bined effective source is to multiply the intensity of the primary source (gray 
body) by the appropriate transmission factor of the filter which is 0.135. This 
factor has been computed from the spectral transmission of the filter via the rela- 
tive energy distribution of 2380 K and the relative visibility function adopted by 
the sixth session of the International Commission on Illumination at Geneva, 

(6) This resolution does not state or imply the value of illumination to be used 
at the test plane during the sensitometric exposure, nor does it place any limita- 
tions on the intensity of the light-source to be used. 

Pyridine as included in Solution A can, in the opinion of various scientific 
groups which have investigated the matter, be obtained commercially of sufficient 
purity to obviate the need for any elaborate precautions and to permit its satis- 
factory inclusion in the formula for Solution A . 


Summary. Proposals of specifications for (1) 35 -mm and 16 -mm raw-stock 
cores, (2) for screen brightness, (3} for rescinding the lantern-slide specifications pre- 
viously included among the SMPE Recommended Practices, and (4) specifications 
for cutting and perforating 35-mm positive raw-stock and negative raw-stock have re- 
ceived initial approval of the SMPE Standards Committee in meeting and final ap- 
proval by letter-ballot of the entire Standards Committee membership. The specifica- 
tions were approved by the Board of Governors on January 24, 1941, and therefore 
are now Recommended Practices of the SMPE. f This approval by the Board of 
Governors is in accord with the new standardization procedure for the Standards 
Committee also adopted by the Board on the same date. 

35-Mm and 16-Mm Raw-Stock Cores. In the Report of the Stand- 
ards Committee published in the January, 1940, issue of the JOURNAL 
were included proposed specifications for 16-mm raw-stock cores, 
after approval by letter-ballot of the SMPE Standards Committee. 
Similar approval by letter-ballot has since been obtained on the 
specifications for 35-mm raw-stock cores. For convenience of refer- 
ence, the two sets of specifications are reproduced on the following 
two pages. 

Screen Brightness. The previous SMPE Recommended Practice 
for screen brightness, published in the March, 1938, issue of the 
JOURNAL indicated 7 to 14 ft-lamberts at the center of the screen, 
when the projector was running with no film in the gate. The re- 
cently approved specifications, contained on a following page, 
changes this range of brightness to 10 i \ ft-lamberts. 

Lantern-Slides. The previous specifications for the mat opening 
of lantern-slides were given in the March, 1938, issue of the JOURNAL 
on p. 255. Letter-ballot of the Standards Committee has recently 
given final approval to rescinding these specifications and to omit 
entirely from the SMPE Recommended Practices any specification 
for lantern-slides. 

Cutting and Perforating 35-Mm Positive Raw-Stock and Negative 
Raw-Stock. Previous specifications for 35-mm negative and positive 
raw-stock were contained in a single drawing published in the March, 
1938, issue of the JOURNAL on p. 261, including the same perforation 
for both negative and positive film. After several years of trial of 




the single perforation, the Standards Committee, by letter-ballot, has 
approved the adoption of different perforations for positive and 
negative raw-stocks. The specifications are given on two following 

If after thirty days from the date of publication of this issue of the 
JOURNAL, no adverse comments are received by the Chairman of the 
Standards Committee from the membership of the Society with re- 
gard to these four items, the specifications described herein will be 
referred to the Board of Governors of the Society for action upon 
them as proposals for either American Standards or Recommended 
Practices. Comments on these proposals are invited from readers of 





D. B. JOY, Chairman 














(Please turn to next page.) 



U. S. M. P. E. 

For 35-mm Film 


January, 1941 


Inch Equivalents 



35.00 + 0.00 
- 0.05 
4.75 == 0.013 
2.794 0.01 
1.85 0.01 
3.40 0.05 
28.17 0.05 
Not > 0.025 
2.08 0.025 
475.00 0.38 

1.378 + 0.000 
- 0.002 
0.1870 0.0005 
0.1100 =*= 0.0004 
0.0730 0.0004 
0.134 0.002 
1.109 0.002 
Not > 0.001 
0.082 0.001 
18.70 0.015 

** Diameter of circle of curvature. 
t L = length of any 100 consecutive perforation intervals. 

* For picture negative and certain special processes. 

These dimensions and tolerances apply to the material immediately 
after cutting and perforating. 

Mar., 1941 



For 35-mn, Film ^ 



1 A 

t C ^ 


1 V ) 




D d 
3 F C 



D -6 




D 6 C 
D C 

1^ ^__ 




/wcA Equivalents 




35.00 + 0.00 
- 0.05 
4.750 * 0.013 
2.794 0.01 
1.98 =t 0.01 
3.40 0.05 
28.17 0.05 
Not > 0.025 
475.00 =*= 0.38 

1.378 + 0.000 
- 0.002 
0.1870 0.0005 
0.1100 =t 0.0004 
0.0780 =t 0.0004 
0.134 =t 0.002 
1.109 0.002 
Not > 0.001 
18.70 0.015 

** L = length of any 100 consecutive perforation intervals. 

* For positive prints and sound recording. 

These dimensions and tolerances apply to the material immediately 
after cutting and perforating. 



[J. S. M. P. E. 

For 35-mrn Film 


January, 1941 






=*= 0.20 
=*= 0.25 
=fc 0.50 

1.968 * 
1.358 =*= 


Recommended Practice 



* 0.30 

0.657 * 
0.157 db 


Bore A to fit freely to hub 25.40 0.1 mm or 1.000 == 0.004 inch diameter. 

Mar., 1941] 



SMPE Recommended Practice 
For 16 -mm Film 


January, 1941 





=t 0.20 
=*= 0.25 
== 0.50 



.020 0.008 
.968 0.010 
.610 =* 0.020 

Recommended Practice 




=*= 0.20 

.657 0.012 
.157 0.008 

Bore A to fit freely to hub 25.40 0.1 mm or 1.000 =*= 0.004 inch diameter. 



For 35-mm Film 

January, 1941 


The brightness at the center of a screen for viewing 35-mm motion 
pictures shall be lOlt ft-lamberts when the projector is running, 
with no film in the gate. 




These theater acoustic recommendations are based upon the 
experience of architects, acoustical and equipment engineers, and 
studio operating personnel in designing, equipping, and maintaining 
theaters. They were prepared after conferences and discussions 
between the Committee and prominent representatives of these 
various groups. Through such cooperation, it is now possible and 
practicable to formulate general principles to guide the acoustic 
design and construction of motion picture theaters. These prin- 
ciples, when applied, will improve sound reproduction and minimize 
or eliminate costly alterations in the completed auditorium. 

In designing a theater auditorium, the architect is interested 
primarily in the usefulness and appearance of the finished structure. 
However, the auditorium shape, and the type, amount, and location 
of the necessary acoustic materials must guide in the construction 
and final appearance. From this point of view, some general con- 
siderations of acoustics will be outlined, and the application of these 
principles explained. In order to avoid misunderstandings, defini- 
tions of certain acoustical terms are given in an appendix. 


The acoustical requirements for good listening conditions in an 
auditorium are that the sound loudness be adequate ; that the com- 
ponents of the complex sound maintain their proper relations; and 
that the successive sounds in fast-moving speech or music be clear 
and distinct, and that the auditorium be free from extraneous noises. 
These fundamental concepts are both necessary and sufficient for 
good listening conditions. 

These proper listening conditions are affected by the following 
physical factors: 

* Presented at the 1940 Fall Meeting at Hollywood, Calif.; received Dec. 20, 
1940. (The Society is not responsible for statements by authors.) 




(7) Size of the room. 
(2) Shape of the room. 

(5) Absorption characteristics of the acoustic materials and their placement 
in the room. 

(4) Extraneous noise level present hi the room. 


As the optimum reverberation time of a room and the proper control 
of reflection effects are the two most important factors in proper 
acoustical design, a brief discussion of these factors follows : 

Optimum Reverberation Time. The desirable reverberation time 
of a room is a function of its size. The effect of moderate reverbera- 
tion is beneficial as the direct sound is reinforced and a desirable 
liveness is produced. In general, a reverberation time of 2 seconds 
(at 500 cycles and under empty hall conditions) should not be ex- 
ceeded. Excessive reverberation causes blurring of speech and 




FIG. 1. Optimum reverberation time for motion picture theaters. 

100, OOO 

rapidly moving staccato music. Where the reverberation time in 
the room is below optimum, an excessive amount of sound energy 
must be radiated, and the resultant sound is unnatural. 

The optimum reverberation period varies with frequency and 
with the size of the room. Fig. 1 gives the optimum reverberation 
time for various sizes of motion picture auditoriums at frequencies 
from 50 to 4000 cycles. 

Reflection Effects. 1 Many theaters still have many major acoustical 
defects due to their shapes and sizes which cause echoes and ob- 
jectionable concentrations by a focusing of the reflected sound. 



These reflections are in numerous cases of greater importance than 
the reverberation time. 2 

When a sound-wave strikes a wall of a theater, its energy is par- 
tially reflected, partially absorbed, and partially transmitted. The 
reflection, for surfaces large in comparison to the wavelength of the 
sound, is analogous to the reflection of light, and the angle of inci- 
dence of the sound is equal to the angle of reflection. The relative 
amounts of energy reflected and absorbed by a material vary with 
the angle of incidence and with the frequency. 

FIG. 2. A typical illustration of optimum acoustical design. 

However, while a doubling in energy is the maximum increase or 
reinforcement which can take place at any one point, the decreased 
level or cancellation effect can be infinite. Consequently, any ab- 
sorption at the point of reflection will tend to decrease the additive 
and subtractive components, and to minimize the modification of the 
characteristics of the direct sound. 

Control of Reflection Effects. For proper sound reflection control 
in an auditorium the acoustic treatment and shape of the walls and 
ceiling must be such as thoroughly to diffuse the reflected sound. 
In other words, the reflected energy received in any auditorium 
location should not come from one particular reflecting area but 


should be contributed by numerous reflecting surfaces. The energy 
from any reflection should be small compared to the total reflected 
sound energy at any point in the auditorium. 

This also provides a uniform logarithmic decay of the reverberant 
sound. 3 Little is gained by attempting to reinforce the direct sound 
by allowing maximum reflection to take place, as any small gain in 
sound energy may be overbalanced by destructive interference 

Optimum Acoustical Design. Two of the most common defects 
of a theater, attributable to poor shape design, are echoes and sound 
concentrations. 3 These, as well as other defects, can be avoided, 
and the optimum characteristics obtained by observing the following 
general rules (see Fig. 2) : 

(1) The cubical contents should be kept to a minimum consistent with the 
number of seats required. 

(2) The auditorium width should be from 50 to 70 per cent of the length, and 
the ceiling height not more than 40 per cent of the length. 

(5) Non-parallel surfaces should be used. 

(4) Convex, rather than concave, walls and ceiling sections should be pro- 
vided. The wall and ceiling surfaces should also otherwise be broken up so as 
to diffuse the sound thoroughly. 

(5) The average absorption per square-foot on the floor and ceiling should 
not be appreciably different from the average absorption per square-foot on 
the side walls. 

(6) Well upholstered seats and ozite-lined carpet hi the aisles should be 

(7) The backstage should be so shaped and so acoustically treated that 
resonant reinforcements of sound will not be reflected into the auditorium to 
distort sound quality. 

It should be pointed out that the design in Fig. 2 is only one 
method of applying the above principles to obtain an ideal set-up. 
The fully convex rear wall and the convex sections on the side walls 
and ceiling are ideal design features, but a design including the three 
convex surfaces on the rear wall or a design as shown by the solid 
lines will also give excellent results. 

These general rules will be explained in greater detail and illustra- 
tions given of good and faulty designs. 


Optimum Volume. Both from an acoustical standpoint and in the 
interests of economy, the cubical content or volume of the auditorium 



should be kept as small as possible, consistent with the required 
seating capacity and proper proportions of length, width, and height. 

In small auditoriums it is possible to obtain optimum reverberation 
conditions without acoustic treatment on the walls and ceiling, 
provided the seats are fully upholstered, the aisles carpeted, and the 
auditorium properly designed to diffuse the sound reflections and to 
provide uniform loudness at every seat in the auditorium. 

Recent acoustical design in which the volume has been held to a 
minimum, without sacrifice of other features, has provided one of 
the most outstanding improvements in general sound quality in the 
theater. From an average of recent trends in design giving satis- 


FIG. 3. Academy Research Council minimum power requirements for 
theaters; minimum recommended amplifier output based upon seating 
capacity of the auditorium. 

factory performance, it has been determined that approximately 125 
cubic-feet per seat is desirable for a medium-size house. 4 This 
figure varies with the seating capacity of the auditorium. 

Minimum Power Requirements for Theaters. The acoustical 
power necessary for proper sound reproduction varies with the volume 
of the auditorium. Minimum power requirements for theaters have 
been established by the Academy Research Council, 5 and are given 
in Fig. 3 in terms of amplifier watts installed. 


Recommended Auditorium Proportions. For some time a propor- 
tion of from 50 to 70 per cent between length and width has been 


considered good. 6 The source of sound is at the end of the small 
dimension. In a room with non-parallel walls or of irregular con- 
tours, the above rule may be applied using the average width and 
length. In those rooms where the length is greater than twice the 
width, the reflections between the side walls cause serious damage 
to the quality. 

As previously explained, excessively high rooms should be avoided; 
that is, the ceiling height should be kept to the minimum consistent 
with the seating capacity and other necessary architectural require- 

Side Walls. Flat parallel walls in auditoriums of all sizes have 
always been a distinct source of trouble from both a vision and 
sound standpoint. In such an auditorium seats installed in the front 
corner sections of the ground floor afford only a very distorted view 
of the picture and are unsatisfactory to the theater patron. From a 
sound standpoint, flat parallel surfaces give rise to disturbing sus- 
tained cross-reflection. A theater with a "fan-shaped" floor plan, 
as shown in Fig. 2, offers decided advantages over rectangular de- 
signs, as better vision is secured, and a basic shape is available which 
is easily accommodated to a side wall design giving proper diffusion 
and control of the reflected sound. 7 

Where it is not practicable to use convex side wall surfaces because 
of economic limitations or difficulties of providing proper lighting, 
the side walls should be broken up into sloping sections. These 
sections should be so angled with respect to the high-frequency horn 
as to reflect the sound well into the side rather than too much into the 
center of the audience. 

Auditorium Ceiling. The ceiling surface should not be parallel 
to the floor, as such a design results in acoustical deficiencies. As 
shown in Fig. 2, the ceiling should be designed to secure desirably 
directed reflections and to eliminate echoes. In all cases every 
effort should be made to avoid ceiling designs which involve domes or 
other concave types of construction which focus the sound into the 
seating area. 

Lighting fixtures constructed with loosely held portions of glass 
or plastic materials should be avoided as such fixtures often rattle, 
being resonant at certain frequencies. 

Auditorium Floor. Well upholstered and heavily redded seats 
should be used in furnishing the auditorium. The aisles and corridors 


should be covered with ozite- lined carpet. This reduces the variation 
in the reverberation time under different audience conditions. 

With upholstered seats and carpeted aisles, it is generally not de- 
sirable to treat the ceiling acoustically. Such treatment results in 
an acoustic condition in which the reverberation time is shorter in 
the vertical direction; that is, shorter between the floor and the 
ceiling than between the side walls and the front and rear walls. 
For optimum acoustical conditions in a room, it is necessary that the 
reverberation die away uniformly in every direction. 

Rear Wall Construction. In the past, rear wall construction has 
often led to serious acoustical deficiencies in the theater. Unbroken 
walls have given rise to reflections into the audience with sufficient 
magnitude and time delay to be audible as echoes. Concave rear 
wall construction with consequent focusing of the reflected sound 
also has adversely affected sound quality. In contrast, convex 
surfaces provide the highest degree of sound dispersion. 

The individual shaping may take different forms but should always 
diffuse the sound. Otherwise the sound reproduction will lack pres- 
ence; that is, the voices of the actors will appear to come from a 
source other than the screen, and the illusion of reality, so important 
to the proper presentation of sound motion pictures, will be lost. 
A modified convex-type rear wall can be utilized to advantage in 
order to conserve seating area and at the same time to eliminate the 
focal difficulties inherent with concave rear wall design. Definite 
breaks in the contours which would have any tendency to form pock- 
ets and produce resonant cavities should be strictly avoided in the 
rear and side walls. 

Illustrations of Bad Acoustical Designs. A typical example of a 
theater with faulty acoustics is shown in Fig. 4. The auditorium 
is rectangular in shape with flat parallel side walls, a flat rear wall, 
and a relatively unbroken ceiling. These walls contribute to multiple 
echoes and a very bad overall acoustical condition. The intelligi- 
bility of the dialog is low. 

Fig. 5 illustrates a side wall section constructed so as to focus the 
sound into the seating area. The concave wall surface adjacent to 
the stage focuses the reflected sound into the center section of the 
seating area, reducing screen presence and intelligibility. (This wall 
section is also shown in Fig. 7, viewed from the rear of the audi- 

By constructing the front portion of the side wall in a convex 



FIG. 8 

FIG; 9 

FIGS. 4-9. Illustrations showing features in theater design, with regard 
to acoustic properties. 

shape as illustrated in Fig. 6, a properly diffusing surface is provided 
and the acoustic defects are avoided. The shape of the wall in Fig. 6 
may be seen by noting the junction of the wall and the ceiling. 
Another example is shown in Fig. 7, where the ceiling is also flat 


and unbroken. In this case it was necessary to tilt the high-frequency 
horn well down into the audience to eliminate reflections from the 
ceiling. This decreased reflection effect provided a decided difference 
in quality between the front and back seating sections. 

A recent trend in rear wall construction is shown in Figs. 8 and 9. 
When designed it was felt that such a tilting of the rear wall would 
reinforce the sound in the back portion of the seating area, giving a 
more uniform loudness level throughout the auditorium. Tests have 
indicated that these constructions are undesirable as this type of 
rear wall tends to concentrate low-frequency sound in rear audience 
areas and contributes to poor sound quality. 

As a result of these experiences, such designs are considered im- 
practicable, and it is recommended that the rear wall generally be 

FIGS. WA and B. Illustrating factors in theater design involved in acoustic 


vertical and of such shape as to diffuse the reflected sound and to 
deliver it back into the audience in random directions. The rear 
wall must always be considered in conjunction with, and never 
separately from, the side wall, ceiling, and floor designs. 

Application of Good Design Principles. A practical example of 
the application of these design principles resulting in a theater with 
exceptionally good acoustics is illustrated in Figs. 1CL4 and B. A 
shows the rear wall and ceiling construction as viewed from the side 
of the auditorium. B is a view from the stage. Fig. IOC shows in 
detail the vertical and horizontal cross-sections of the auditorium 
and the three different rear wall cross-sections. 

This particular auditorium has been constructed on one of the 
studio lots and seats approximately 700 persons. However, in a 


commercial application the rows would be closer together, so this 
theater may be considered to have a seating capacity of about 1000. 

Although the principles of good design are fulfilled, certain features 
depart from the ideal given in Fig. 2. Fortunately, such deviations 
may be made without serious penalty. The side walls, while of 
straight sections, have been broken up to give a series of differently 
sloped small reflecting surfaces with the direction of reflection well 
into the side of the seating area. The same construction has been 
used on the ceiling. The rear wall surfaces consist of flat areas but 
the overall contour is convex, and the walls are sectioned in irregular 
sizes and at different angles so that each section contributes only a 
small reflection in any particular direction. It should be noted that 
all sections of the rear wall are vertical and do not tilt toward the 
seating area. 

Backstage Acoustical Treatment. In the reproduction of sound in 
the theater one of the most critical acoustical requirements is ade- 
quate backstage draping. In the design of the loud speaker and its 
associated unit every effort is made to minimize the amount of sound 
radiated from the back of the loud speaker. However, a certain 
amount of ' 'leakage" radiation takes place and must be adequately 
absorbed. It is realized that efforts along this line have been made 
by various architects, but in general the absorption provided has not 
been sufficient. 

Although the rear wall of the stage is often covered with rock- wool 
blankets, this material, when applied directly to a brick wall or other 
hard surface, has been inadequate. In other cases the absorption 
material has been set out from the wall on furring strips with the 
absorption material applied to wire mesh stretched across the strips 
to increase the low-frequency absorption properties. This is far 
more effective than the former method but still does not reduce the 
backstage reverberation time to the necessary negligible amount. 

It has been observed that in a large number of cases this treatment 
is finished with a white surface, and for proper picture projection it 
is suggested that a black covering be used instead to avoid light 
reflections back to the screen. 

Considering the type of backstage absorption treatment normally 
provided it has been necessary for the theater owners completely to 
cover the back of the screen with ozite, with the exception of the space 
occupied by the loud speakers. 

Suitably located drapes around and above the loud speaker provide 





FIG. IOC. Cross-section, plan, and elevation of theater shown in 
Figs. 10(4) and (). 

an efficient means of absorbing undesirable backstage sound reflec- 
tions. However, this type of treatment is an additional expense 
for the exhibitor and some form of speaker draping and screen 
masking should be combined in the initial design. 

In the design of the stage of the auditorium, two factors should be 
borne in mind. First, for proper viewing and listening conditions 
the first row of seats should be at least 20 feet from the screen, where 
the screen is not more than 16 feet in width. For wider screens the 
first row of seats should be back an additional 15 inches for each foot 
of screen width over 16 feet. Second, the stage floor should be 
shaped to give an unobstructed view from the front seats, and the 
stage area should be covered with a rug or other sound-absorbing 
material to eliminate reflections directly from the loud speaker into 
the seating area. 



The absorption characteristics of commercial acoustic materials 
vary widely. Some materials also show a pronounced absorption 
peak in their characteristics. These materials, while highly efficient 
for certain purposes (absorption of typewriter clicks, for instance) 
may be detrimental in a theater as their use produces a non-uniform 

The materials selected should show a smooth absorption character- 
istic and adequate absorptivity at the low frequencies. For this 
reason care should be taken in selecting the material to be used in an 
auditorium in order that the desired overall effect be secured. 

Modern acoustic materials are of such varied design and con- 
struction that it is possible for the architect to apply the proper 
sound-absorbing material and at the same time produce almost any 
artistic effect desired. 

Low- Frequency Absorption. Due to the fact that many acoustic 
materials are deficient in absorption at the low frequencies, it is often 
necessary to use these materials in a particular way to effect sufficient 
low-frequency absorption in the auditorium. To obtain this ab- 
sorption at the extreme low frequencies the diaphragm action of 
the walls should be utilized. This action may be explained as 
follows: Where a wall is not rigidly supported, as in the case of a 
furred-out wall, it will act as a diaphragm and vibrate at certain 
frequencies, depending upon the dimension of the wall, the mass per 
unit surface, and the manner of wall support. This diaphragm action 
may either reinforce or absorb the sound near the particular frequency 
at which the wall resonates. If the wall has little internal resistance 
it will continue to vibrate after the original sound has died down 
and will re-radiate energy into the room, reinforcing the original 

In contrast to this, if the material has high internal resistance to 
vibration (as do masonite, celotex, and similar acoustic materials) 
the vibrations will die down faster than the reverberation time in the 
room and no "hang-over" effect takes place. To avoid a boomy 
house (which has excessive response in the region from 60 to 300 
cycles) it is often advisable to make use of this diaphragm action to 
absorb these low registers. 

In the practical application of such design, the furred-out walls 
should be made into different sized sections by irregular bracing of 
the wall itself. Such construction provides wall sections of different 



dimensions which absorb the low frequencies through a wider band 
and avoid dips or peaks in the response characteristics. 

N on- Symmetrical Absorption Areas. 9 The most recent design 
considerations have proved the efficacy of using completely non- 
symmetrical small areas of absorption materials as contrasted with 
the earlier use of large treated surfaces. This non-symmetrical 
arrangement tends to maintain the long indirect path reflections and 
reduce the formation of surface patterns resulting in unsatisfactory 
sound conditions. 







FIG. 11. Optimum amounts of sound absorption material for motion picture 


The amount of sound absorption material necessary in theaters of 
various sizes is given in Fig. 11. This chart is useful for design 


A theater, to have good acoustics, should have its walls insulated 
against the transmission of outside noise into the auditorium. 9 The 
transmission of sound is of two kinds: (1) air borne, and (2) struc- 

Small openings around doors, windows, through port-holes, etc., 


transmit sound readily. For this reason all the joists between 
walls, doors, and windows which lead outside should be made as tight 
as possible. Transmission of sound through the building structure 
(such as the noise from vibrating machinery) can be minimized by 
using double-wall and double-floor construction, where required, and 
by separating all vibrating machinery from their supporting structure 
with vibration isolating materials such as cork or rubber. Massive 
walls are not always necessary to obtain sufficient sound insulation. 
A double wall of fairly light construction will give good sound in- 
sulation provided the two walls are not closely coupled mechanically 
by nails or rigid close members. 

A large part of the transmitted noise often comes from the projec- 
tion room. For this reason, as much fireproof acoustic material as 
possible should be placed on the inside walls and ceiling of the room. 
Since much of the projection room noise is radiated through open 
portholes or portholes with glass windows, these too should be treated. 

It is also recommended that the air-conditioning system be operated 
at a low noise level, by employing a- large-volume, slow-velocity 


In summarizing these theater acoustic recommendations reference 
is again made to Figs. 2 and 10, and the foregoing principles presented 
in outline form. The essential design features are: 

(1) A minimum volume consistent with the required seating capacity and 
proper auditorium proportions. 

(2) An auditorium width of from 50 to 70 per cent of the length and an 
auditorium ceiling height of not more than 40 per cent of the length. 

(5) The use of non-parallel surfaces; in particular, the floor should not be 
parallel to any ceiling section or opposite side wall sections parallel. 

(4) The use of convex rather than concave surfaces. In addition the wall 
and ceiling surfaces should otherwise be broken up so as to thoroughly diffuse the 

(5) Auditorium absorption characteristics to provide the same rate of sound 
decay in a vertical as in a horizontal direction from side to side or from back to 
front walls. 

(6) Heavily upholstered seats and ozite-lined carpet in the aisles. 

(7) Backstage treatment giving a negligible amount of reflected or re-radiated 
sound from the backstage into the auditorium. 

(5) A heavily carpeted proscenium designed for good viewing conditions from 
the front seating section. 

(9) Auditorium walls with sufficient sound insulation material to prevent 
extraneous noise entering the auditorium. 


(10) The projection room acoustically treated with fireproof material and 
projection ports equipped with acoustic baffles. 

(11) All equipment subject to vibration and hum such as arc generators, 
voltage regulators, lighting control equipment, etc., acoustically isolated from the 

(12) Air-conditioning equipment of a high- volume, low air- velocity type with 
air ducts provided with acoustic baffles. 

Long narrow auditoriums, high ceilings, excessively long and narrow balcony 
overhangs, concave focusing surfaces, and large unbroken reflecting areas should 
always be avoided as acoustical faults will always result from their use. 

If these recommendations are followed, the resulting auditorium 
will give sound (as reproduced on a modern two-way equipment) 
with high intelligibility, warm, natural screen presence, good balance 
between high and low frequencies, uniform loudness level throughout 
the auditorium and the proper relative balance between high-level 
music passages and low-level, intimate dialog scenes. 


Acoustic Absorptivity. The acoustic absorptivity of a surface is equal to one 
minus the reflectivity of that surface. 

Acoustic Reflectivity. The acoustic reflectivity of a surface which is not an 
original source, is the ratio of the rate of flow of sound energy reflected from the 
surface, on the side of incidence, to the incident rate of flow. 

Diffuse Sound. Sound is said to be in a perfectly diffuse state when, in the 
region considered, the energy density, averaged over portions of the region large 
compared to the wavelength, is uniform and when all directions of energy flux 
at all parts of the region are equally probable. 

Echo. An echo is a wave which has been reflected or otherwise returned with 
sufficient magnitude and delay to be perceived in some manner as a wave distinct 
from that directly transmitted. 

Echo, Flutter A. flutter echo is a rapid succession of reflected pulses resulting 
from a single initial pulse. If the flutter echo is periodic and if the frequency is 
in the audible range it is called a musical echo. 

Echo, Multiple. A multiple echo is a succession of separately distinguishable 
echoes from a single source. 

Mean Free Path. The mean free path for sound waves in an enclosure is the 
average distance sound travels in the enclosure between successive reflections. 

Rate of Decay. The rate of decay of sound energy density is the time rate at 
which the sound energy density is decreasing at a given point and at a given time. 
The practical unit is the decibel per second. 

Reverberation. Reverberation is the persistence of sound, due to repeated re- 

* From "Acoustic Terminology," American Tentative Standard, Bull. Z24.1, 
1936, American Standards Association. 


Reverberation Time. The reverberation time for a given frequency is the time 
required for the average sound energy density, initially in a steady state, to 
decrease, after the source is stopped, to one-millionth of its initial value. The 
unit is the second. Thus the time required for a sound to decay 60 db is the re- 
verberation time. 

Sabins. The sabin is a unit of equivalent absorption and is equal to the 
equivalent absorption of one square-foot of a surface of unit absorptivity, i. e., 
of one square-foot of surface which absorbs all incident sound energy. 


1 MORRIS, P. M.: "Some Aspects of the Theories of Acoustics," J. Acoust. 
Soc. Amer., XI (July, 1939). 

2 BOLT, R. : "Normal Modes of Vibration in Room Acoustics," /. Acoust. Soc. 
Amer., XI (Oct., 1939). 

3 KNUDSEN, V. O.: "Some Practical Aspects of Architectural Acoustics," J. 
Acoust. Soc. Amer., XI (April, 1940). 

4 POTWIN, C. C. : "The Control of Sound in Theaters and Preview Rooms," 
/. Soc. Mot. Pict. Eng. t XXXV (August, 1940), p. 111. 

6 "The Theater Standardization Activities of the Academy Research Council," 
Acad. Research Council Tech. Bulletin (May 22, 1940). 

6 POTWIN, C. C.: "Theater Acoustics Today," Better Theaters Magazine 
(May 29-Oct. 16, 1937). 

7 RETTINGER, M.: "Motion Picture Theater Developments," /. Soc. Mot. 
Pict. Eng., XXXIV (May, 1940), p. 524. 

8 POTWIN, C. C., AND MAXFIELD, J. P.: "A Modern Concept of Acoustical 
Design," /. Acoust. Soc. Amer., XI (July, 1939). 

9 VOLKMANN, J. E., AND MoRRiCAL, K. C. : "Fundamentals of Theater Acous. 
tics," Internal. Proj., XV (May, 1940). 





Summary. This report includes a brief outline of the reasons for the formation of 
this Sub- Committee, the problems under consideration, and the manner of approach- 
ing a solution of these problems. 

In addition, tests and acoustical measurements as previously made are described 
and certain conclusions drawn. 

This report of the Academy Research Council's Sub-Committee on 
Acoustical Characteristics is intended primarily as a progress report 
on the Sub-Committee's program and will include only such con- 
clusions as seem justified at the present time. 

However, a brief outline of the reasons for the formation of the 
Sub-Committee, the problems under consideration, and the manner 
of approaching a solution of these problems, will be of interest to the 
members of the Society as the results of the work will eventually be 
of major benefit to all those engaged in recording and reproducing 
sound motion pictures. 

This Sub-Committee was formed as a result of the work of, and 
functions under, the Council's Committee on Theater Sound Stand- 
ardization, and so our discussion will be preceded by a brief resume 
of some phases of that Committee's work. 

In the early days of sound recording and up to the advent of the 
commercial two-way loud speaker system, the greatest effort in sound 
motion pictures had been toward improvement in recording equip- 
ment and technic. At that time it was felt that this phase of sound 
motion pictures had advanced more rapidly than the complementary 
reproducing equipment and technic. After the two-way loud speaker 
systems had been installed in many theaters in the United States, a 
decided trend toward uniformity of reproduced sound was expected, 

* Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received Dec. 20, 
1940. (The Society is not responsible for statements by authors.) 
** Chairman. 

284 J. DURST [j. s. M. P. E. 

but as a result of several field inspection trips by members of the Re- 
search Council it was evident that instead of a tendency toward uni- 
formity there was a divergence in quality from theater to theater as 
well as between the products of different producers in the same 
theater. This was due to differences in the electrical characteristics 
of reproducing equipment as adjusted in the theater, bad acoustical 
conditions, and differences in recording characteristics. 

As a result of this trend the Theater Sound Standardization Com- 
mittee was formed to study the problem and to work toward uni- 
formity in sound reproduction. Realizing that both art and science 
have played a most important part in sound motion pictures, there 
was no thought of introducing a degree of uniformity approaching 
monotony or lack of the effect of creative effort in sound as repro- 
duced in the theater. On the contrary, it was felt that this uniform- 
ity was necessary in order that the producers and the sound de- 
partments, fore-armed with a knowledge of the reproducing char- 
acteristic, could obtain the utmost in naturalness and showmanship. 
Nor was there any thought that this standardization program was 
one which would be completed in a short time, but a program origi- 
nally intended to cover a period of years was undertaken. 

As a first and most important step, the Research Council Standard 
Electrical Characteristics were established for a majority of two-way 
reproducing systems. To establish these Standard Electrical Char- 
acteristics, a great number of listening tests were conducted in many 
theaters in the Southern California area, on each type of equipment 
under consideration. Theaters were chosen which were considered 
to have "average" acoustical conditions in the sense that acoustical 
deficiencies were not obviously present as shown by previous ex- 
perience in listening to all types of products reproduced at previews 
and regular showings. 

The Research Council Theater Sound Test-Reel (ASTR-2), con- 
sisting of typical current product from the major studios was made 
up by the Theater Standardization Committee. By listening to this 
material in the various theaters, data upon which the Standard Elec- 
trical Characteristics were based were obtained and the characteris- 
tics eventually established. 

In consideration of the manner in which these Standard Character- 
istics were obtained, through listening tests in a great number of 
acoustically good theaters, it was felt that the acoustical effect of any 
one auditorium was minimized as the electrical characteristics speci- 


fied were set up on a basis which actually included the average effect 
of theater acoustics. 

The Committee fully realizes the incompleteness of such a speci- 
fication since sound pictures are reproduced also in theaters having 
poor acoustics. It was also recognized that the specification of an 
acoustical characteristic would eventually give the complete and most 
eminently satisfactory answer to this particular phase of the problem. 

The Research Council felt that steps should be taken to investi- 
gate acoustical measurements and this Sub-Committee was ap- 
pointed to consider this problem. 

In the opinion of the Sub-Committee, the problem resolved itself 
into six distinct sub-problems : 

(7) The selection of satisfactory portable measuring equipment. 

(2) The establishment of a suitable sound source making acoustical mea- 
surements reproducible within satisfactory tolerances under the most severe 
acoustical 'conditions. 

(5) The establishment of suitable calibrating and measuring technic most 
applicable for these purposes. 

(4) The study of acoustical response characteristics of review rooms. 

(5) The determination of the efficacy of the equipment and the technic of its 
use by making an extended series of measurements to gather data on acoustical 
response conditions in a number of average theaters. 

(6) The correlation of listening tests and acoustic measurements in theaters 
and review rooms. 

It is readily apparent that the response characteristic of review 
rooms and theaters must be made to correspond before the studio 
personnel can definitely know, by listening in their projection room, 
how their product will sound in the field. In consequence, this cor- 
relation between review room and theater is an important and vital 
step in the standardization program of the Research Council. 

It is definitely to the interest of this Committee in viewing the 
overall picture to recommend a technic which will be most practicable 
in the field. The considerations along these lines are quite varied 
but it is our aim to recommend a method and procedure which will, 
by virtue of its economy, be an asset to its users. It was obvious from 
the start that little progress could be made until such an equipment 
and such a method of measurement which would give reproducible 
results were available. 

Columbia Pictures Corporation kindly made available to the Sub- 
Committee one of their studio review rooms for making preliminary 
tests and measurements. The curves in Fig. 1 show the results of 



LT. S. M. P. E 

-5 - 

-10 - 
-15 - 
-20 - 

FIG. 1. Average. 
FIG. 2. Position 1. 
FIG. 3. Position 2. 
(ERPI , RCA , GR ) 

our initial efforts. These are the averages of measurements shown in 
Figs. 2, 3, 4, and 5 for selected individual locations. As indicated, 
three types of measuring equipment were employed: namely, the 
ERPI high-speed level recorder with a pressure type microphone, 
the RCA sound level meter using a velocity type microphone, and a 
General Radio noise-meter with a crystal microphone. A warble 


film having an 8-cycle rate and a =*= 5 per cent modulation factor was 
used as a sound-source. 

While the variation in the results was definitely discouraging, work 
was carried on exploring the many possibilities of different sound- 
sources such as random emission noise, multitone, and various types 
of warble tones. 

By employing the same calibrating equipment and an identical 
sound-source, calibrations of all the measuring equipments were 
made by this group, at all times employing the same measurement 
technic. ERPI made available their equipment for these calibra- 
tions. Check tests made under the same acoustical environment 
showed a very gratifying improvement. The deviations which pre- 
viously approached 10 db were reduced to the order of 2 db. 

The Altec Service Corporation conducted a very extensive study 
of warble tones and published their findings in the February, 1940, 
issue of the JOURNAL. With their cooperation a new warble film was 
recorded having the characteristics of warble rate and modulation 
which their research disclosed as being most suitable to this type of 

The upper curves of Fig. 6 show the results of later measurements 
made with improved sound-sources and recalibrated equipments. 
They represent the averages of the curves of Figs. 7, 8, 9, and 10. 
For comparison purposes, Fig. 1 has been reproduced on the same 

It is interesting to note the consistency of the results as well as the 
general smoothing out of the measured characteristics. This latter 
effect is due largely to the changes made in the sound-source, and the 
resulting smoother curve more nearly corresponds with what the ear 
hears. From the standpoint of accomplishment it is of interest to 
make direct comparisons not only of the averages but also of the in- 
dividual positions; for example, it will be observed that the extreme 
"raggedness" originally measured in position 1 has now become a 
relatively smooth characteristic. The original measurements made 
in position 1 could not be correlated with listening tests. Yet these 
later measurements were made with the same measuring equipments 
except that they had been accurately calibrated and the rate as well 
as the percentage of modulation of the warble film had been changed. 
Repeated tests made at substantial intervals proved that these 
measurements could be readily duplicated within practical toler- 
ances. Furthermore, these measurements permitted a satisfactory 



[]. S. M. P. E. 

FIG. 4. Position 3. 
FIG. 5. Position 4. 
FIG. 6. Average. 
(ERPI , RCA , GR ) 

degree of correlation with listening tests. Direct comparisons were 
made by listening to both the warble tone and the Academy Theater 
Test-Reel in the various positions used for measurements, and it was 
agreed that the two tests gave positive indication that we had meas- 
ured the differences which were detectable by ear. 

Mar., 1941] 



For further verification of this, a test was made in a medium-size 
theater with another group of observers. In this case, the acoustical 
response characteristic, as shown Fig. 11, was set without the knowl- 



FIG. 7. Position 1. 

FIG. 8. Position 2. 

FIG. 9. Position 3. 

(ERPI , RCA - -, GR ) 

edge of any of the observers. The observers were permitted to 
listen only to the Academy Theater Test-Reel and to indicate what 
they heard in terms of relative frequency response. Without ex- 
ception, it was agreed that there was an excess of energy at 150 cycles, 
a lack of energy from 500 to 700 cycles, and that the high end was 



[J. S. M. P. E. 

held flat to approximately 4000 cycles and sharply attenuated above 
this frequency. 

With this evidence it was felt that our present measuring proce- 
dure not only lends itself to correlation with listening but also pro- 
vides a definite means of diagnosis of improperly adjusted character- 



.10 - 

FIG. 10. Position 4. 

FIG. 11. Undesirable acoustic characteristic confirmed by listen- 
ing tests (average of five positions). 
(ERPI , RCA , GR ) 

The Committee has solicited the cooperation of the two major 
theater service organizations in the interests of obtaining acoustical 
response characteristics in several hundred theaters. The measure- 
ment procedure is given in an Appendix. 

With the data obtained from these tests and the assistance of the 
Theater Standardization Committee, it is believed that an optimum 
accoustical characteristic may be established. It must be borne in 
mind, however, that listening tests must be the final and basic cri- 
terion for the determination of theater system adjustments. The 
entire program of this Sub-Committee has been predicated on the 

Mar., 1941] 



theory that the development of a scientific tool as an adjunct to list- 
ening will permit better standardization of acoustical response char- 
acteristics throughout the industry. 

We wish to take this opportunity to acknowledge the helpful co- 
operation of Altec Service Corporation, Columbia Studio, Electrical 
Research Products, Inc., General Radio Company, International 
Projector Corporation, Lansing Manufacturing Company, Para- 
mount Studio, RCA Manufacturing Company, Samuel Goldwyn 
Studio, and Warner Brothers Studio, for helpful suggestions, the use 
of studio facilities, testing equipment, and sound -sources. 



(Having Modern Two-Way Speaker Systems] 

(1) For these response measurements 6 positions are chosen: 5 positions in 
the auditorium and 1 stage position, as indicated on Fig. 12. If the house is 
symmetrical, positions should be located in the same half of the house. No posi- 
tion should be closer than 10 feet to any reflecting surface such as a wall or pillar. 
In making these response measurements use the microphone mounted on a tripod 
and keep all observers at least 10 feet away from the microphone during the run. 



FIG. 12. Showing positions for response measurements. 

(2) For the stage position, locate the microphone about 10 feet in front of the 
screen and about 18 inches from the theater centerline. The microphone should 
be approximately at the height of the top of the low-frequency horn or baffle. 
The Academy Research Council Warble Film (No. 6490} is projected at normal 
fader setting for the theater and the readings recorded on the data sheet under 










5 S 

















" c 

























5 i 



^ S 


c j 









, 5 





















i 5 




.-__ 1 1 

U 5 ;- 

. 1 1 


*z Q 

i i 

o L 


St o 

j S 







.* S 


> Q 

o Q 










1 s ^ 









-. Q 











- -s" 





35 S? 

- Q 



" | EC. 




^o J 



g" " 

n S 





















n n N n 


lilt" 2 - 

Mar., 1941] 



"Stage Position" together with the height of the microphone and its distance from 
the screen. 

(5) At position 1 (Fig. 12) locate the microphone on a seat at the average ear 
height of a seated observer. Record the row and seat number in the "Position 1" 
column of the data sheet (Fig. 13). Using the same machine and fader setting, 
project the warble film as before, and record the sound level readings in the 



FIG. 14. Cross-section required for data sheet. 

"Position 1" column of the data sheet. Repeat this procedure in positions 2, 3, 
4, and 5. Record the row and seat number of each postion and the sound level 
eadings in the appropriate column of the data sheet. 

(4) Record on the data sheet the information requested to aid in evaluating 
the acoustical measurements. In the space provided sketch the vertical and hori- 
zontal cross-sections of the theaters as illustrated in Fig. 14. 



In July, 1916, a group of 26 men, farseeing and prominent engi- 
neers, who were then pioneering the problems of a growing industry, 
got together in Washington and organized what they called the So- 
ciety of Motion Picture Engineers. Mr. C. Francis Jenkins, engineer 
and inventor, residing in Washington, acted as temporary chairman. 
At this meeting, which may be regarded as an organization meeting, 
a Constitution and By-Laws were drawn up and adopted, in which 
the aims of the Society were set forth as "the advancement in the 
theory and practice of motion picture engineering and the allied arts 
and sciences, the standardization of the mechanisms and practices 
employed therein, and the maintenance of a high professional stand- 
ing among its members." 

Immediately the new group began to hold meetings to which were 
attracted quite a number of persons interested in the then relatively 
young art of motion picture engineering. 

Mr. Jenkins, after the completion of the organization, was elected 
President of the Society, with Mr. D. G. Bell of Chicago, and Mr. E. 
M. Porter of New York as Vice-Presidents, and with Mr. E. K. Gil- 
lette of New York and Mr. Paul Brockett of Washington as Secretary 
and Treasurer, respectively. Other notable persons on the member- 
ship list at that time were Messrs. Carl Akeley, Max Mayer, and 
Herman Kellner. 

By the time of the first formal meeting, which was held in October, 
1916, the membership had grown considerably, numbering about 40 
persons. Of that small group, we still have with us, as members of 
the Society, Messrs. A. S. Victor, F. H. Richardson, and W. C. Kunz- 
mann. All the other members of that time have since either died or 
otherwise dropped out of our membership. 

* Reprinted from the Modern Theater Section of Box Office, Oct. 12. 1940, p. 56. 
** President, Society of Motion Picture Engineers, 1939-40. 


The technical activities of the new Society gained a good start at 
the first formal meeting in 1916 through the formation of the follow- 
ing committees : 

Committee on Cameras and Perforations 

Committee on Picture Theater Equipment 

Committee on Motion Picture Electrical Devices 

Committee on Optics 

The dues in those days were quite interesting. Charter member- 
ships in the Society were divided as follows : 

Pioneer Member $250 . 00 

Honorary Member 100 . 00 

Active Member 25 . 00 

Associate Member 25.00 

Regular memberships in the Society (that is, not Charter member- 
ships) were established in the following grades : 

Active Membership Entrance Fee $25.00 

Annual Dues 10.00 

Associate Membership Entrance Fee 25.00 

Annual Dues 5.00 

Those responsible for the organization of the Society wisely made 
its aims and purposes rather broad, providing for the development of 
its activities in many and diverse fields of an industry destined to 
grow to enormous size and to cover in its activities a wide range of 
scientific and artistic endeavor. There is little doubt, however, that 
the one thing uppermost in the minds of the men who formed this 
Society, the one thing they considered of prime importance, and of 
immediate necessity for the welfare of the industry, was the standardi- 
zation of materials, mechanisms, and practices. As a matter of 
fact, at the first meeting, although only one paper was presented, 
that by H. D. Hubbard of the U. S. Bureau of Standards, the sub- 
ject of that paper consisted entirely of a discussion of standardiza- 
tion, and the necessity of standardization in the motion picture in- 

At the second meeting, at which time Mr. Jenkins was chosen 
President, the President's address was a plea for standardization, 
and two or three papers on the program dealt directly with the stand- 
ardization of motion picture film and mechanism. Although no 

296 E. A. WILLIFORD [j. S. M. P. E. 

formal "Standards Committee" had been formed, it was from this 
beginning that the present standardization activities of the Society 
have grown. 

From the beginning, regular meetings of the Society were held in 
the spring and fall of each year. These meetings were more in the 
nature of conventions at which groups of papers were presented and 
discussed, similar to those the Society now holds semi-annually. It 
is interesting that the method of conducting conventions has not 
changed in all these twenty-four years; and it is also interesting to 
note that one of our present members, our Convention Vice-President, 
Mr. W. C. Kunzmann, was not only present at the first formal 
meeting of the Society but has attended every Convention that the 
Society has ever held. 

The papers presented at each meeting were collected into groups 
and published in two volumes a year. These volumes were known 
as the "Transactions of the Society of Motion Picture Engineers." 
Some of the early issues were rather thin, but each year, up to and 
including 1929, showed a steady growth. The first Transactions 
contained only one paper the address by Mr. Jenkins. The second 
Transactions contained three papers. The Transactions of 1928 
embodied 1222 pages of printed material. 

In the meantime, the Society had steadily grown, although at 
first rather slowly. In 1928 the total membership was 320, with 185 
Actives and 135 Associates. It will be noted that in these days the 
Active membership exceeded the Associate. 

In 1929, sound came into the picture, and, with the addition of 
electrical engineers, radio engineers, and other kinds of engineers to 
the industry, the membership of the Society took a big jump. In 
1931, the membership had leaped to 765, almost equally divided be- 
tween the Active and Associate grades. 

About this time it became recognized that great loss of value and 
great inconvenience resulted from concentrating the publication of 
papers presented at meetings into two volumes of the Transactions 
a year. It had become increasingly important for engineers of the 
industry to receive their technical information promptly and regu- 
larly. For that reason, in 1930, through the insistent efforts of Dr. 
Loyd A. Jones and Mr. John I. Crab tree, who was then President, 
the Society was fortunate in being able to enlist the financial assis- 
tance of various companies of the industry for the purpose of expand- 
ing the technical scope of the Society and putting the Transactions 

Mar., 1941] BETTER PROJECTION 297 

on a monthly basis. A fairly large group of producing companies 
and equipment manufacturers subsidized the Society to the extent 
of almost $10,000. The name Transactions was discontinued, 
and in January, 1930, the first monthly JOURNAL was issued. 

Up to this time, engineers had been regarded by many motion 
picture executives as being very much apart from the industry; and, 
in fact, at a banquet tendered to the Society, meeting in Hollywood 
in 1928, by the Academy of Motion Pictures Arts and Sciences, one 
of the outstanding executives of a producing company admitted that 
prior to that occasion he had not even heard of the Society of Motion 
Picture Engineers. The motion picture technician had at last been 
given the spotlight, and, since that time, his importance and value to 
the motion picture industry have been more and more deeply realized. 
It can not be disputed that the success of the motion picture industry, 
as we know it today, lies to an enormous extent in the hands of the 

The Society of Motion Picture Engineers contributed in no small 
way to the successful accomplishment of the transition from the 
silent picture to the sound picture, and the Transactions contained 
the first published accounts of the first systems of recording and re- 
producing sound. 

During the first year of the issuance of the JOURNAL (1930) Dr. L. 
A. Jones acted as temporary Editor. The clerical and administra- 
tive work of the Society up to this time (1930) had been attended to 
by officers acting without remuneration, in their homes or in the 
offices of their firms. The steady growth of the Society and its suc- 
cess in coordinating the technical activities of the industry can in 
very large measure be attributed to the sacrifices made and the in- 
conveniences endured by those officers conducting their Society work 
under what were at best inadequate facilities. 

Just as prior to 1930 it had long been the dream of the governors 
to put the Transactions on a monthly basis, so, with the growth 
of the industry and of the Society, and with the undue burdens placed 
upon the men who fostered his growth, the Board began to give 
thought to the establishment of permanent headquarters for the So- 
ciety, and the employment of a man for the purpose of relieving the 
officers of much of these burdens. 

Accordingly, permanent headquarters of the Society were estab- 
lished at 33 W. 42nd Street, New York, N. Y., with Mr. Sylvan 
Harris as Editor of the JOURNAL and Office Manager, in which ca- 

298 E. A. WILLIFORD [j. s. M. P. E 

pacity lie has been acting for the past ten years. In 1934 the office 
was moved to the Hotel Pennsylvania. 

The scope of the Society's technical activities grew very rapidly; 
the number of technical committees was increased to embrace all 
phases of sound engineering, acoustics, and many of the new arts 
brought into the motion picture field. Additional technical com- 
mittees were established to cover phases of the industry that had 
heretofore been neglected, such as projection practice. The list of 
committees, as constituted today, presents a most imposing picture 
of the breadth of the Society's technical activities. The technical 
committees at present functioning are : 


Exchange Practice 

Laboratory Practice 

Non-Theatrical Equipment 

Preservation of Film 




Studio Lighting 


Theater Engineering (includ- 
ing projection practice and 
theater design) 

In addition, the Society holds representation in the American 
Documentation Institute, the Sectional Committee on Motion Pic- 
tures of the American Standards Association, the Sectional Com- 
mittee on Photography of the American Standards Association, and 
the Inter-Society Color Council. 

Attending the economic fiasco in 1929, the membership suffered a 
severe loss, dropping from a total of 765 to 560. Broad revisions of 
the Society's membership policies and alterations of the membership 
dues were then made so that it would be more easily possible for 
members of the industry to become members of the Society. A 
third grade of membership was created, known as Fellow membership, 
and the dues of the other grades were revised to the following figures : 

Fellow $20.00 

Active 15.00 

Associate 10.00 

Admission fees were eliminated. With these changes, and with 
the gradual reconciliation of the people to the new economic situation, 

Mar., 1941] BETTER PROJECTION 299 

and with the help of a vigorous membership campaign, under the 
guidance of Mr. E. R. Geib, the membership again began to increase. 
Subsequent revision of the dues to the present figures in force, given 
below, added considerably to the growth, and despite unsettled world 
conditions during the past several years, there has been a steady in- 
crease of membership since the low point of 1933 up to the present. 
At this time the Society's membership is the largest it has ever been, 
totaling almost 1400 members. This membership includes all sorts 
of persons engaged in the motion picture industry and its allied arts 
and sciences, and many others who make motion pictures their avoca- 

It is interesting to note these various groups of persons who are 
members of the Society: 

Laboratory technicians 

Photographic chemists 



Equipment manufacturers 

Equipment designers 



Theater managers 

Sound recordists 

Illuminating engineers 



Exchange managers 

Film editors 


Acoustical engineers 

Amateur cinematographers 

There are technical experts of all the important research labora- 
tories and other engineering organizations throughout the entire 
world, exclusive of the manufacturing, producing, and exhibiting 
branch of the industry, and the roll of members includes the names of 
many distinguished scientists and engineers, in addition to the 
honorary members of the Society. There is only one honorary mem- 
ber of the Society at the present time, namely, Mr. Thomas Armat 
of Washington, D. C., who was one of the pioneers of the motion 
picture. Others who have been honorary members of the Society 
but whose names, at their deaths, have been placed upon the So- 
ciety's Honor Roll, are as follows : 

300 E. A. WlLLIFORD [J. S. M. P. E. 

Louis Aime Augustin Le Prince 
William Friese-Greene 
Thomas Alva Edison 
George Eastman 
Frederic Eugene Ives 
Jeane Acme Le Roy 
C. Francis Jenkins 
Eugene Augustin Lauste 
William Kennedy Laurie Dickson 

Probably one of the most important activities of the Society is its 
standardization program. The standards that have been established 
by the Society include dimensional specifications of film and perfora- 
tions and sprockets, the relation between emulsion and sound-track 
positions in scanned areas, reels, projection lenses, splices, screen 
sizes, projection room lay-out, sensitometry, photographic density, 
screen brightness, and motion picture nomenclature. Many of these 
standards have passed from the status of motion picture industry 
standards to American national standards; and some have actually 
become international in scope, having been approved by the Inter- 
national Standards Association just prior to the present war in 
Europe. The Society's interest includes all sizes of motion picture 
film and equipment, in addition to the customary 35-mm professional 

As an important technical contribution to the industry, several 
years ago the Society, through the Projection Practice Committee, 
designed and made available to the motion picture industry a sound 
test-film and a visual test-film, designed for the purpose of checking 
the operation of equipment in the motion picture theater. These 
films have since been used very extensively throughout the world 
and to quite a large extent by various departments of the United 
States Government. Many companies, and a number of govern- 
mental departments, have based their purchases of projection and 
sound reproducing equipment upon performances as checked by 
these SMPE test-films. 

All the technical activities of the Society are presented, described, 
and discussed at the conventions of the Society, which are held twice 
a year. The sessions usually last four days and are attended by 
hundreds of members and their friends and relatives. At the semi- 
annual banquet, with a customary attendance of several hundred, 
each visitor, whether delegate or guest, carries away with him the 
memory of a great occasion. Technical papers, followed by discus- 

Mar., 1941] BETTER PROJECTION 301 

sions of the papers by the members, demonstrations of newly de- 
veloped apparatus and methods, open-forum discussions of the prob- 
lems of the industry, and exhibitions of equipment are technical 
features of the conventions. These are supplemented by an interest- 
ing program of social and entertainment features. 

The headquarters of the Society are located in New York, where 
members of the Society and their friends are always welcome. It is 
the seat of the editorial activities, and the avenue through which the 
various functions of the Society are administered. All members are 
welcome to visit the General Office at any time, and to take advan- 
tage of such privileges as it may be able to provide in the form of 
service to the members. The office is located on the mezzanine of 
the Hotel Pennsylvania, Seventh Avenue and 34th Street, New 
York, N. Y. 

Local Sections of the Society have been established for the purpose 
of enabling the members in the three great motion picture centers of 
the United States to conduct and attend meetings in their own dis- 
tricts. The three sections are the Atlantic Coast Section, which 
meets regularly at New York or nearby; the Mid- West Section, 
which operates for the benefit of members residing in and about 
Chicago ; and the Pacific Coast Section, which meets in Hollywood. 

Membership in the Society is open to anyone interested in the 
motion picture art. The two grades of membership make it possible 
for anyone to participate in the Society, whether a novice at the art, 
or an eminent authority in any of the many fields of endeavor in the 
motion picture industry, whether interested in motion pictures as an 
avocation, or because of business or professional connections with 
the industry. The fees are low, having been reduced to the lowest 
level at which it is still possible for the Society to serve the industry 



Min. of M. P. Number of Annual 

Grade Age Work References Dues 

Active 25 3 3 Fellows or Actives $15.00 

Associate 18 1 Fellow or Active 7.50 

The Society's office will gladly furnish, upon request, further in- 
formation covering membership in the Society, and will be glad to 
extend its f acilities to anyone in the motion picture industry, whether 
a member or not, in search of technical or other relevant informa- 


Summary. A line microphone is a microphone consisting of a large number of 
small tubes with the open ends, as pick-up points, equally spaced along a line and the 
other ends connected by means of a common junction to a transducer element for con- 
verting the sound vibrations which converge upon the junction into the corresponding 
electrical variations. Several types of line microphones with the useful directivity 
along the line axis are described as follows: a simple line, a line with progressive 
delay, and two lines with progressive delay and a pressure gradient element. 

The directional characteristic of a microphone is an expression of 
the variation of behavior of the microphone with respect to direction. 
The solid angle over which sound is received without appreciable 
attenuation relative to the maximum response characterizes a direc- 
tive sound-collecting system. The effective solid angle of the di- 
rectional characteristic determines the ratio of direct to generally 
reflected or other undesirable sounds. The shape and magnitude of 
the solid angle of the directional characteristic determines the ratio 
of direct to generally reflected sound. One important requirement 
is a directional characteristic which is independent of the frequency. 
A system which does not possess this characteristic will introduce 
frequency discrimination in both the desired and undesired sounds. 

The preferable directional characteristic of a microphone for a 
particular application will depend upon the pick-up problem. For 
example, the bidirectional velocity microphone has been found to 
be very useful for overcoming excessive reverberation, eliminating 
undesirable sounds, and as a tool for obtaining a "correct balance" 
of the received sound. The unidirectional microphone has been 
found to be particularly useful where the desired sounds originate 
in front of the microphone and the undesired sounds at the rear of 
the microphone. These microphones have demonstrated the useful- 
ness of directional microphones in sound motion picture recording, 

* Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received Sep- 
tember 30, 1940. (The Society is not responsible for statements by authors.) 
** RCA Manufacturing Co., Camden, N. J. 



radio broadcast, phonograph recording, and sound reinforcing sys- 
tems. It has been found that directional sound-collecting systems 
are more effective than nondirectional systems in problems connected 
with sound collection, such as balance and reverberation in music, 
long distance pick-up of speech, and feedback difficulties in sound 
reinforcing. In view of this demonstrated usefulness, further con- 
sideration has been given and efforts directed toward the develop- 
ment of other directional systems. It is the purpose of this paper 
to describe a number of directional microphones of the line type. 

A line microphone is a microphone consisting of a large number of 
small tubes with the open ends, as pick-up points, equally spaced 
along a line, and the other ends connected by means of a common 
junction to a transducer element for converting the sound vibrations 
which converge upon the junction into the corresponding electrical 
variations. In the line systems to be considered the transducer will 
be a ribbon element located in a magnetic field and terminated in an 
acoustic resistance. Under these conditions the outputs of the pipes 
may be added vectorially. By suitable impedance terminations other 
transducer elements may be used to obtain the same type of opera- 
tion, that is, so that the outputs may be added vectorially. 

Simple Line Microphone. The simple line microphone consists 
of a number of small pipes, with the open ends, as pick-up points, 
equally spaced on a line and the other ends joined at a common 
junction, the lengths of the pipes decreasing in equal steps (Fig. 1). 
A ribbon element, in a magnetic field, connected to the common 
junction and terminated in an acoustic resistance in the form of a 
long damped pipe, is used for transforming the acoustical vibrations 
at the ribbon into the corresponding electrical variations. 

The pressure contribution by any element n at the common junc- 
tion of the line may be expressed as 

^ n = B n ^^^( x n - x n coa VX (1) 

where/ = Frequency, in cycles per second. 
t = Time, in seconds. 

x n = Distance of the element n from the center of the line, in centimeters. 
6 = Angle between the axis of the line and the incident sound. 
B n = Amplitude of the pressure, in dynes per square-centimeter, due to 
the element n. 

In the case of a uniform line with the strength a constant, the re- 
sultant when all the vectors are in phase is B n l where / is the length of 
the line. 



[J. S. M. P. E. 




FIG. 1. Simple line microphone and the directional char- 
acteristics as a function of the length of the line and the 

The ratio Rg of the response for the angle 6 to the response for 6 = 

The absolute value of the expression on the right hand of equation 
2 is given by 


H "" ". 

sin*(/ - Jcosfl) 

J (J - / cos 0) 


Mar., 1941] 



The directional characteristics of the simple line microphone for 
various ratios of length of the line to the wavelength are shown in 
Fig. 1. These characteristics are surfaces of revolution about the 
line as an axis. 

Line Microphone with Progressive Delay. Referring to Fig. 1, it 
will be seen that the length of the line must be of the order of several 
wavelengths to obtain a directional characteristic with a relatively 
small effective solid angle. Practically, the simple line type of 
microphone becomes quite long if high directivity is desired at the 


FIG. 2. Line microphone with progressive delay and the directional char- 
acteristics, for a time delay equal to one-quarter the length of the line, as a 
function of the length of the line and the wavelength. 

lower part of the audible spectrum. However, it is possible, by 
introducing into the pipes a delay which becomes progressively greater 
with the length of the individual pipes, to effect an increase in direc- 
tivity; that is, a decrease in the effective solid angle of pick-up over 
that obtained with the same length of simple line. As in the case of 
Fig. 1, the microphone of Fig. 2 consists of a number of small pipes 
with the open ends as pick-up points equally spaced along on a line 
and the other ends joined at a common junction, with the addition 
that there is inserted a delay which is proportional to the distance 

306 H. F. OLSON [J. S. M. P. E. 

from the end of the line or the pick-up point nearest to the common 

The ratio RQ of the response for the angle 6 to the response for 
6 = is 



where d is the path length of the delay introduced for the point 
farthest removed from the common junction. 

sin- [I - I cos + d] 
? [I - I cos 6 + d] 


The directional characteristic of the line microphone, with pro- 
gressive delay for various ratios of the length of the line to the wave- 
length, is shown in Fig. 2. A measure of the value of a line with 
progressive delay as compared to a simple line may be obtained by 
comparing Figs. 1 and 2. It will be seen that the same directional 
characteristic can be obtained with a line of shorter length by intro- 
ducing appropriate delay. In the case of a delay comparable to the 
wavelength, loss in sensitivity occurs. 

Two-Line Microphone with Progressive Delay and a Pressure- 
Gradient Element. This microphone consists of two lines of the type 
shown in Fig. 2 arranged so that the ribbon element measures the 
difference in pressure generated in the two lines (Fig. 3). The centers 
of the two lines are separated by a distance D. In the line nearest 
the ribbon element, a bend of length D is inserted between the 
junction and the ribbon element. 

To show the action of the pressure-gradient element, assume that 
the small pipes are removed and the two junctions, at Ji and Jz, 
Fig. 3, are used as the pick-up points. These two pick-up points 
are then separated by a distance D. The difference between the 
forces on the two sides of the ribbon, assuming the mass reactance of 
the ribbon system to be large compared to the resistance of the 
damped pipes, may be expressed as 

F M = A cos (2rf/) sin ( rLf cos \ (7) 

where A is a constant, including the pressure of the impinging sound- 
wave and dimensions of the microphone. 

Mar., 1941] 



If D is small compared to the wavelength, equation 7 becomes 

F M = A ? cos (27T/0 cos 6 


Equation 8 shows that the force available for driving the ribbon is 
proportional to the frequency and the cosine of the angle 0. 

Employing a mass-controlled ribbon of mass m r , the velocity is 
given by 

FIG. 3. Pressure gradient line microphone consisting of two lines with 
progressive delay and a pressure gradient element and the directional char- 
acteristics for a time delay equal to one-quarter the length of the line, as a 
function of the length of the line and the wavelength. 

Where c is the velocity of sound, in centimeters per second. The 
velocity x is independent of the frequency for a constant sound 
pressure, and as a consequence the ratio of the generated electro- 
motive force to the pressure in the sound-wave will be independent 
of the frequency. 

In the above discussion the junctions are assumed to be non- 
directional. The directional characteristics of the individual lines 
are given by equation 6. Therefore the expression for the directional 
characteristics of the combination of these lines with a pressure- 
gradient element is the product of equations 6 and 9. The directional 
characteristic is given by 

308 H. F. OLSON 

sin- (/ - I cos e + d) 
R = - cos e (10) 

* (I - I cos 6 + d) 

The directional characteristics of the microphone of Fig. 3, for 
various ratios of the length of the line to the wavelength for a delay 
of one-half the length of the line, combined with a pressure-gradient 
element, are shown by the graphs of Fig. 3. A measure of the 
improvement in directivity obtainable from a line with progressive 
delay and a pressure-gradient element may be obtained by comparing 
Figs. 1 and 3. Employing these expedients, approximately the same 
directivity may be obtained with a line of one-quarter the length 
necessary for a simple line. 

Conclusion. Directional microphones employing lines of various 
types have been described, the directional characteristics of which 
indicate considerable variation with frequency. Experience has 
shown that a directional characteristic which varies with frequency 
is undesirable because frequency discrimination is introduced in the 
direct sound for sources of sound removed from the axis. In addi- 
tion, the response to reflected sound is a function of the frequency, 
and the reverberation characteristics of the received sound is thereby 

Experiments upon directional systems have indicated that a 
microphone with a small solid angle of pick-up would be extremely 
useful in certain applications such as in sound motion picture record- 
ing, in television pick-up, in symphony and stage pick-up, in sound 
broadcasting, and in sound reinforcing. The line type of microphone 
discussed in this paper appears to be the logical solution from the 
standpoint of size and portability. However, the directional char- 
acteristics must be independent of the frequency This result 1 can 
be accomplished by employing a number of separate lines each cover- 
ing a certain portion of the frequency range. 


1 ANDERSON, L. J. : "A Line Type of Microphone for Speech Pick-Up, " /. 
Soc. Mot. Pic. Eng., XXXVI (Feb., 1941), p. 309. 


During the Conventions of the Society, Symposiums on new motion picture appara- 
tus are held, in which various manufacturers of equipment describe and demonstrate 
their new products and developments. Some of this equipment is described in the 
following pages; the remainder will be published in subsequent issues of the Journal. 



The reduction of the line type of microphone 1 to a practical form is compli- 
cated by the fact that in order to secure useful directional characteristics at low 
frequencies, the microphone becomes extremely large and cumbersome. Even the 
use of delay in the pipe system and the use of a pressure-gradient pick-up device 
necessitates the use of lines whose length approaches one-half wavelength at the 
lowest frequency involved. Bearing in mind the fact that quite satisfactory speech 
pick-up can be obtained with a system whose response is limited to a band width 
of 150 to 5000 cycles, and, that in order to produce a microphone of reasonable 
portability the length could not exceed 6 feet, an investigation was made to see 
what might be accomplished in the way of directional characteristics. 

For a simple line type, 6 feet long, with maximum response along the axis, the 
angle at 150 cycles will be about 160 degrees, which is rather broad, and at 5000 
cycles will become far less than 30 degrees. Such variations in directional charac- 
teristics are undesirable since they represent differing response characteristics for 
points off the axis. Unless it is desired to complicate the structure considerably, 
the angle of response at 150 cycles must be accepted as a starting point. How- 
ever, the extreme narrowness at higher frequencies may be eliminated by con- 
structing the microphone in such a way that the effective length of the line be- 
comes shorter with increasing frequency. In practice the theoretical ideal of 
having the reduction proportional to the frequency is approximated by dividing 
the microphone into three or more sections, with either electrical or acoustical 
filtering means. Further analysis of the problem indicates that dividing the 
response range into three bands will produce directional characteristics which do 
not vary too widely. The bands selected were, roughly, 200-750 cycles, 750-2100 
cycles, and 2100-5000 cycles. The length of the section that is to function in each 
band is then chosen so that the product of the average frequency and the length 
is approximately constant from section to section. (See Figs. 1 and 2.) 

The low-frequency response may be limited to the desired value of 150 cycles 

*Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received October 
15, 1940. (The Society is not responsible for statements by authors.) 
**RCA Manufacturing Co., Camden, N. J. 



FIG. 1. (Right) Longitudinal view. 
FIG. 2. (Left) Side view. 

by any one of several means. However, the simplest way in which the desired low- 
frequency cut-off may be obtained is through the use of a diaphragm-type pres- 
sure microphone in which the normal low-frequency response is obtained by a 
resonant circuit consisting of the case volume and an air mass in a tube connecting 
the case volume to the outside. By closing the tube, the stiffness of the moving 
system and case may be used to secure the low-frequency cut-off. While the cut- 
off so obtained is not rapid, it is sufficiently so to prevent accentuation of the 
lower frequencies due to the decreasing directivity. The upper cut-off frequency 
for the low and mid-frequency pipes is obtained by inserting in each pipe close to 
the pick-up unit a plug with a small hole. The inertance so introduced, hi combi- 
nation with the pipe impedance and the volume between the microphone dia- 
phragm and the pipe termination, provides the desired cut-off. The cut-off at 
5000 cycles for the high-frequency pipes is obtained by providing a small compli- 
ance between the microphone diaphragm and the ends of the pipes. No inertances 
are inserted in the high-frequency pipes. 

Since only one microphone functions at the high-frequency end, two micro- 
phones in the mid-frequency range, and three microphones in the low-frequency 

Mar., 1941] 



range, it is necessary to provide the individual tubes of the high and mid- fre- 
quency sections with suitable horns. 

The high-frequency section is placed as close to the associated microphone unit 
as possible in order to avoid attenuation in the pipes. It consists of four pipes, 
each branching into three smaller pipes (totaling an area equal to that of the larger 
pipe). The resulting twelve openings are spaced 1 inch apart, making a line 11 
inches long, which provides pick-up points sufficiently close for all practical pur- 
poses. These pipes are fitted with double-flare horns for the purpose mentioned 
above. The horn section nearest the pipe has a cut-off frequency of 750 cycles 
and a gain of 2 : 1 so as to function in combination with the mid-frequency horns. 
The second section of the horn was given a cut-off frequency of 2100 cycles and a 
gain of 5 : 1 to compensate for losses in the coupling to the microphone and in the 
microphone proper, due to the absence of diffraction effects, which are normally 
present. The mid-frequency section consists of six pipes, with openings spaced 2 
inches apart, approximately 2 inches from the nearest high-frequency pipe. 
These tubes are fitted with horns having a cut-off frequency of 750 cycles and a 

FIG. 3. Frequency response on axis. 

gain of 2:1. Because of physical limitations imposed by the use of branched 
pipes and horns in the high-frequency section, the mid-frequency section contains 
the shortest tube of the entire assembly. The low-frequency section consists of a 
bundle of nine pipes, terminating at spacings to bring the overall length of the 
microphone to approximately 6 feet. The gains obtained with the horns specified 
are somewhat greater than necessary, but are advantageously used to secure in- 
creased sensitivity from the microphone. In the experimental model the desired 
response was obtained when the output transformers of the high and mid-fre- 
quency sections had about one-third the turn-ratio of the one used in the low- 
frequency section. Figs. 1 and 2 give a much clearer idea of the general arrange- 

Summarizing the operation of the microphone, we have the following: between 
200 and 750 cycles, all the pipes and microphones are functioning, and the effec- 
tive length of the line is approximately 6 feet. Between 700 and 2100 cycles, only 
the high and mid-frequency sections are functioning, by virtue of the inertance 
introduced into the low-frequency pipes. The. loss of sensitivity due to the drop- 
ping of the one pick-up unit is compensated for by the 750-cycle sections of the 
horns on the high and mid-frequency lines. From 2100-5000 cycles, only the 
high-frequency pipes are in operation, because of the inertances inserted in low 



and mid-frequency pipes. Above 5000 cycles, the response is attenuated as de- 
sired by the proper choice of compliance between the microphone diaphragm and 
the pipe ends. 

The results of tests made on the experimental model are shown in Figs. 3 and 
4, the data for which were obtained out of doors with the microphone approxi- 
mately 28 feet from the sound-source. The agreement with theoretical predictions 
is generally good except at the lower frequencies where the directional characteris- 
tics are somewhat narrower than expected. 


FIG. 4. Directional characteristics. 

Observations regarding the performance under operating conditions are as yet 
relatively meager, but it appears that microphones of this type should offer a 
powerful new tool for the pick-up of speech under adverse conditions. 


1 OLSON, H. F.: "Line Microphones," J. Soc. Mot. Pic. Eng., XXXVI (Mar., 
1941), p. 302. 


MR, HOPPER: We have had some experience with a similar type of micro- 
phone, developed by the Bell Laboratories some years ago. It has similar direc- 
tional properties. Two models were built, one five feet long, and one ten feet 
long. The 5-ft model was directional to 200 cycles, and the other to an octave 
lower. Using the microphone indoors on sets and stages, due to the fact that low- 
frequency troubles are usually predominant on sets, there was not sufficient di- 
rectivity at low frequencies and the instrument did not function too satisfactorily. 
The rather large size of the device proved objectionable on a set congested with 
the usual picture equipment. For stage production use the response characteristic 


should be rather like that of the average type of microphone normally used on the 
set, so that recordings made on either microphone will be sufficiently alike to per- 
mit direct intercutting of takes. Outdoors the instrument works more success- 
fully because we are not so concerned with low-frequency reverberation or reflec- 
tion from surrounding objects. 

Have you had any experience with the microphone in wind? We found that 
the pipe structure used was rather susceptible to wind noise. 

MR. RETTINGER: I have not had a great deal of experience with the micro- 
phone outdoors. I calibrated it outdoors, but made the tests when there was no 
wind, usually in the mornings. 

MR. SKINNER: Is the size of the pipes very important? Could they be made 
smaller? Is a definite diameter required for each pipe? 

MR. RETTINGER: There is no limit to the size of the tube that can be used, 
except that when the tube becomes too small there is a great deal of attenuation 
due to viscosity, which is negligible in this case, except perhaps at 5000 cycles. 
These tubes are l /* inch in diameter; the other tubes were 3 /i6 inch. 

MR. SKINNER: By using more tubes could you not get a smoother response? 

MR. RETTINGER: Yes, but there is a limit to the number of tubes that can be 
used without making the diameter of the microphone too great. The weight 
would also increase. The combined areas of the tubes come together into the 
small diameter of the microphone diaphragm, and the large change in area repre- 
sents an acoustic mismatch, which, of course, is not desirable. 

MR. WOLFE: In connection with Mr. Hopper's remarks, we unfortunately 
have had practically no experience with this microphone to date. It came into 
our hands only a few days ago and we do not know what problems will be en- 
countered. We were aware of the problems that were encountered with the Bell 
Laboratories device at the time it was developed, and I feel reasonably sure that 
Dr. Olson and Mr. Anderson have taken them into account, so I believe that the 
variation from the directional characteristic when used indoors will not be as 
great as was previously experienced. 

The quality of this microphone is not as good as that of the microphone we 
commonly use today in recording for motion pictures. But it does seem, if this 
directional frequency characteristic can be maintained, that under difficult pick- 
up conditions the overall frequency response, when all reflections and reverbera- 
tions are taken into account, may be definitely better and more nearly like the 
quality received by a normal microphone under good pick-up conditions. 

MR. TASKER: I might mention that some of us still use concentrators in really 
difficult spots not often, but occasionally. They are certainly bad. 

Are delay circuits used in the longer tubes of this microphone? 

MR. RETTINGER: No. The tubes could have been made shorter, but it was 
not felt that delay tubes would be of sufficient advantage in this microphone. 

MR. CRABTREE: To get the maximum efficiency is it necessary to sight the 
microphone like a rifle? 

MR. RETTINGER: It is helpful. The microphone has a reception angle of 30 
degrees, in which sound is received with uniform intensity. 

MR. CRABTREE: Does it have a telescopic sight? What is the distance at 
which it works most efficiently? 

MR. RETTINGER: It is operated at a little greater distance than the ordinary 


microphone, due to the narrow angle of pick-up. To cover a wider area it must be 
moved farther from the sound-source. 

MR. CRABTREE : Suppose you sighted the rifle at a particular person in a large 
audience, would you get a better response than with the use of a concentrator? 

MR. RETTINGER: The microphone is not so directional as to require an actual 
sight on the object. Most of the energy of speech centers around 500 cycles, at 
which frequency the microphone is still directional, but not so highly so that you 
would not be able to make a good guess at the direction in which to point it. 

MR. TASKER: Since the patterns for the various frequency ranges are similar, 
broadening somewhat at the lower frequencies, the one for 2000 cycles may be 
taken as typical. It is apparent if the subject is anywhere within approximately 
30 degrees, there is substantially maximum response at that frequency. The 
change in response is about two db. If, however, the subject were standing much 
to either side of the 30-degree angle, the amount of sound at 2000 cycles would be 
very low indeed. It is simply necessary to sight this angle of 30 degrees at the per- 
son who is speaking. 

MR. CRABTREE: What is maximum range of satisfactory operation? 

MR. RETTINGER: The microphone can be operated at the same distance as 
any other microphone. As a matter of fact, its sensitivity is higher than that of 
any other type. That was found out also by the Bell Laboratories people when 
they made their first investigations, using a 6 18 A microphone. The sensitivity 
at all frequencies went up 5 db. 

MR WOLFE : These directional characteristics were measured at a distance of 
27 feet from the source. In normal practice we endeavor to keep the microphone 
within five feet of the speaker. That is not always possible, but it is a reasonable 
distance for operating our present recording types of microphone. In public ad- 
dress work we put them much closer to the subject 

With this type of microphone we would normally expect to work at a distance 
that would be perhaps three times as great, or 15 to 25 or 30 feet from the subject. 

Dr. Olson has described 1 other microphones having similar directional charac- 
teristics but developed for a somewhat different purpose. They were designed to 
have a better frequency characteristic than this microphone, because the inten- 
tion was to use them, in some cases, for pick-up of symphony orchestras and simi- 
lar material. Such a microphone was built and used in the Metropolitan Opera 
House in New York, where it was located on the balcony at a distance of 100 to 
150 feet from the orchestra. 

It is a little difficult to define the distance at which the microphone can be oper- 
ated, as that depends entirely upon the circumstances. We can, however, say 
that its normal operating distance is two or three times that of the regular types 
of microphone that have been used. This microphone, or one similar to it, was 
used at the Republican and Democratic Conventions this year by NBC. They 
had very good results in picking up speech from the floor. The microphone must 
frequently have been 50 to 75 feet from the speaker. 



The editors present for convenient reference a list of articles dealing with subjects 
cognate to motion picture engineering published in a number of selected journals. 
Photostatic copies may be obtained from the Library of Congress, Washington, D. C., 
or from the New York Public Library, New York, N. Y. Micro copies of articles 
in magazines that are available may be obtained from the Bibliofilm Service, Depart- 
ment of Agriculture, Washington, D. C. 

American Cinematographer 

22 (February, 1941), No. 2 
Fantasound Disney's New Sound System (pp. 58-59, 80, 

81) W. STULL 

Cut with Script and Camera Says Director Rene Clair (pp. 

60, 82) W. STULL 

New Multiplex Lamp (pp. 61, 82) T. GAUDIO 

Educational Screen 

20 (January, 1941), No. 1 
Motion Pictures Not for Theaters (pp. 15-17), Pt. 23 A. E. KROWS 


14 (January, 1941), No. 1 

Sound in Motion Pictures (Review 1927-1939) (pp. 17-19, 

73-76) N. LEVINSON 

International Projectionist 

15 (December, 1940), No. 12 

"Fantasound" Soundheads and Amplifiers (pp. 18-20) C. S. ASHCRAFT 


22 (October, 1940), No. 10 

Vergleichende Messung der Wirkung verschiedener Schutz- 
benhandlungsmethoden auf Filmbildschichten und deren 
Trager (Comparative Measurements of the Effect of 
Different Kinds of Protective Treatment on the Emulsion 
and in the Film Base) (pp. 141-143) W. FERMAZIN 

Der AFIFA Pruffilm 40/16T fur Schmaltonfilmappa- 
raturen (AFIFA Test Film 40/16T for Substandards 
Sound Cameras) (pp. 143-145) H. ECKELMANN 

22 (November, 1940), No. 11 

Agfacolor Der deutsche Farbenfilm (Agfacolor the Ger- 
man Color Film) (pp. 151-154) H. BECK 
Lichtstarke und relative Offnung von Kine-Objektiven 



(Light Intensity and Relative Aperture of Motion Picture 

Lenses) (pp. 154-157) A. KOCHS 

Kinematograph Weekly 

287 (January 9, 1941), No. 1760 

Revolutionary Matte Process, Applicable to Monochrome 
and Colour (pp. 21-22) 

Motion Picture Herald 

142 (February 1, 1941), No. 5 

Television Brought to Theater by RCA Large Screen Show- 
ing (pp. 30-32) 

Photographische Industrie 

38 (November 27, 1940), No. 48 

Neue Entwickler fur Kinofilm (New Developer for Motion 
Picture Film) (pp. 721-722) 

38 (December 11, 1940), No. 50 

Rechnerische Bestimmung von Regenerierungslosungen fiir 
Maschinen-Entwickler (Mathematical Determination of 
Regeneration Solutions for Machine Developers) (pp. 





Program and Facilities 

E. HUSE, President 

E. A. WILLIFORD, Past-President 

H. GRIFFIN, Executive Vice-P resident 

W. C. KUNZMANN, Convention Vice-P resident 

A. C. DOWNES, Editorial Vice-President 

G. A. BLAIR, Chairman, Local Arrangements 

S. HARRIS, Chairman, Papers Committee 

J. HABBR, Chairman, Publicity Committee 

J. FRANK, JR., Chairman, Membership Committee 

H. F. HEIDEGGER, Chairman, Convention Projection 

Reception and Local Arrangements 





G. A. BLAIR, Chairman 
C. E. K. MEES 

Registration and Information 

W. C. KUNZMANN, Chairman 

Hotel and Transportation 

F. E. ALTMAN, Chairman 

Publicity Committee 

J. HABER, Chairman 






318 SPRING CONVENTION [j. s. M. P. E. 

Banquet and Dance 

I. L. NIXON, Chairman 




Ladies' Reception Committee 

MRS. C. M. TUTTLE, Hostess 






H. F. HEIDEGGER, Chairman 





Officers and Members Rochester Projectionists Local No. 253 


The headquarters of the Convention will be the Sagamore Hotel, where excellent 
accommodations and moderate rates are assured. A reception parlor will be pro- 
vided as headquarters for the Ladies' Committee. 

Hotel reservation cards will be mailed to the members of the Society early in 
April. They should be filled out and mailed immediately to the Sagamore Hotel 
so that suitable accommodations may be reserved, subject to cancellation if 
unable to attend the convention. 

The following European-plan day rates are extended by the Sagamore Hotel to 
Society members and guests attending the Convention (all rooms are outside 
rooms with bath) : 

Room for one person $3 . 00 to $5 . 00 
Room for two persons, double bed 4 . 50 to 6 . 00 

Room for two persons, twin beds 6.00 to 7.00 

Suite accommodations, one to four persons 12.00 and up 

The following hotel garage rates will be available to SMPE delegates and guests 
who motor to the Convention: 24-hr, inside parking, 75 ; outside parking (daily), 

The colorful Sagamore Room on the main floor of the Hotel offers special break- 
fast, luncheon, and dinner menus at moderate prices. 

Golfing privileges at several Rochester country clubs may be arranged for 
either by the hotel management or at the SMPE registration headquarters. 

Mar,, 1941] SPRING CONVENTION 319 


Convention registration and information headquarters will be located on the 
Sagamore Hotel roof, adjacent to the Glass House, where all technical sessions and 
symposiums will be held. 

Members and guests attending the Convention will be expected to register, and 
so help to defray the Convention expenses. Convention badges and identifica- 
tion cards will be provided for admittance to all regular and special sessions during 
the Convention. The identification card will also be honored through the 
courtesy of Loew's Theaters, Inc., at Loew's Rochester Theater and, through the 
courtesy of Monroe Amusements, Inc., at the Palace, Regent, and Century Theaters. 

Group visits to various plants in Rochester and vicinity may be arranged at 
the Registration Desk. 


All the technical sessions of the Convention will be held in the Glass House 
on the roof of the Sagamore Hotel with the exception of Wednesday evening, as 
described below. Members should note that the banquet, which at past con- 
ventions has always been held on Wednesday evening, this time has been sched- 
uled for Tuesday evening to permit holding a special meeting on Wednesday 
evening at the Eastman Theater. 

Wednesday, May 7th, will be devoted to a joint meeting of the Acoustical So- 
ciety of America and the SMPE, consisting of a symposium of papers by engineers 
of the Bell Telephone Laboratories in the morning and afternoon. In the evening 
a demonstration of stereophonic sound will be given by the Bell Telephone Labora- 
tories at the Eastman Theater. 


The usual Informal Get-Together Luncheon for members, their families, and 
guests will be held in the Starlight Room on the hotel roof on Monday, May 5th, 
at 12:30 P. M.: short addresses by prominent speakers; names to be announced 
later. Luncheon and banquet tickets should be procured when registering. 

The 48th Semi-Annual Banquet and Dance will be held in the Starlight Room- 
on the hotel roof on Tuesday evening, May 6th, at 7:30 p. M. : music and enter- 
tainment. Banquet tickets should be procured and tables reserved at registra- 
tion headquarters by noon of Tuesday, May 6th. 


Mrs. C. M. Tuttle, Convention Hostess, and members of her Committee are 
arranging a very attractive program of entertainment for the ladies attending the 
Convention. A reception parlor will be provided for the use of the Committee 
during the Convention. 



Monday, May 5th 

9:00 a. m. Sagamore Hotel Roof 

9:30 a. m.-12:00 Glass House, Hotel Roof 

Technical session 
12:30 p. m. Starlight Room, Hotel Roof 

Get-Together Luncheon for members, their families, and 
guests. Brief addresses by several prominent speakers 
2:00 p. m. Glass House, Hotel Roof 

Technical session 
8:00 p. m. Glass House, Hotel Roof 

Technical session 

Tuesday, May 6th 
9:00 a. m. Sagamore Hotel Roof 

9:30 a. m. Glass House, Hotel Roof 

Society Business 

Technical session 
12:30 p. m. Luncheon period 

Program for this afternoon to be announced later 
7:30 p. m. Starlight Room, Hotel Roof 

Semi- Annual Banquet and Dance of the SMPE: addresses 

and entertainment: music, dancing, and entertainment 

Wednesday, May 7th 
10:00 a. m. Glass House, Hotel Roof 

Technical session 

12:30 p. m. Luncheon period 

2:00 p. m. Glass House, Hotel Roof 

Stereophonic sound papers session 
8:00 p. m. Eastman Theater 

Stereophonic sound demonstration for the SMPE Conven- 
tion and invited groups. Admission only by SMPE 
identification card, or special invitation card 

Thursday, May 8th 
10:00 a. m. Glass House, Hotel Roof 

Technical session 

12:30 p. m. Luncheon period 

2:00 p. m. Glass House, Hotel Roof 

Technical and business session 



Convention Vice- President 



At a meeting held at the Hotel Pennsylvania, New York, on February 19th, 
attention was given to the current motion picture activities in relation to the 
National Defense Program. Two papers, illustrated by 16-mm motion pictures, 
were presented as follows: 

Civilian Aspects of the Defense Program 


Mr. Arch E. Mercy, Special Assistant, Advisory Commission to the 
Council on National Defense, Washington, D. C. 

The Production of Training Films 


Lt. Col. M. E. Gillette, Signal Corps, Photographic Section, Officer in 
Charge of Production of Motion Picture Training Films for the U. S. 

The meeting was very well attended and aroused considerable interest in the 
government motion picture activities. A lively question and answer period 
followed the presentations. 


The proposed amendments to the By-Laws given below have been approved 
by the Board of Governors on the dates indicated. In accordance with the 
procedure for amending the By-Laws, these proposals are published herewith, 
and are to be submitted to the membership of the Society at the business meeting 
to be held at the Spring Convention at Rochester. 

By-Law IV, Sec. 4(b). To the list of standing committees of the Society 
appointed by the Engineering Vice-President shall be added: 
Committee on Process Photography 
Committee on Preservation of Film 

The latter committee has been functioning for a number of years, but has not 
been specifically mentioned in the By-Law. (Approved by the Board of Govern- 
ors, Oct. 20, 1940.) 

By-Law VII, Sec. 1. In the procedure for nominating and electing officers of 
the Society, it is at present stipulated that: 

The Secretary shall then notify these candidates of their nomination in order 
of nominations and request their consent to run for office. 

The proposal is to delete the latter part of this sentence so that it will read: 



The Secretary shall then notify these candidates of their nomination. (Ap- 
proved by the Board of Governors Oct. 20, 1940.) 

By-Law XI, Sec. 6. This By-Law outlines the procedure for nominating and 
electing the officers and managers of the local sections of the Society. The 
present wording is as follows: 

The remainder of the procedure shall be in accordance with the procedure 
specified for the election of officers of the General Society as described in By-Law 
VII, Section l(a), the word Manager being substituted for the word Governor. 

Instead of allowing the procedure to be implicit, by the substitution of the 
word Manager for the word Governor, it was felt that it would be better and 
more accurate if the procedure were written out in full, following, as far as possible, 
the procedure for electing officers of the General Society. In addition to the 
advantage of being explicit, this procedure must differ slightly in wording in 
view of the different compositions of the Board of Governors (for the General 
Society) and the Board of Managers (for the Local Sections). The proposed 
wording is as follows: 

The Chairman of the Section shall then notify these candidates of then* nomi- 
nation. From the list of acceptances, not more than two names for each vacancy 
shall be selected by the Board of Managers and placed on a letter ballot. A 
blank space shall be provided on this letter ballot under each office, in which 
space the names of any Active, Fellow, or Honorary members other than those 
suggested by the Board of Managers may be voted for. The balloting shall 
then take place. 

The ballot shall then be enclosed in a blank envelope which is enclosed in an 
outer envelope bearing the local Secretary-Treasurer's address and a space for 
the member's name and address. One of these shall be mailed to each Active, 
Fellow, and Honorary member of the Society, residing in the geographical area 
covered by the Section, not less than forty days in advance of the annual Fall 

The voter shall then indicate on the ballot one choice for each office, seal the 
ballot in the blank envelope, place this in the envelope addressed to the Secretary- 
Treasurer, sign his name and address on the letter, and mail it in accordance 
with the instructions printed on the ballot. No marks of any kind except those 
above prescribed shall be placed upon the ballots or envelopes. 

The sealed envelope shall be delivered by the Secretary-Treasurer to his 
Board of Managers at a duly called meeting. The Board of Managers shall 
then examine the return envelopes, open and count the ballots, and announce 
the results of the election. 

The newly elected officers and managers shall take office on the January 1st 
following their election. 


At a recent meeting of the Admissions Committee at the General Office of the 
Society, the following applicants for membership were admitted to the Associate 

Mar., 1941] 



542 South Broadway, 
Los Angeles, Calif. 
BALL, W. S. 

1256 Howard St., 

San Francisco, Calif. 

Photo and Sound, Inc., 
153 Kearny St., 

San Francisco, Calif. 

1816 Silverwood Terrace, 

Los Angeles, Calif. 
BOYD, J. M. 

2013 S. Vermont Ave., 

Los Angeles, Calif. 

1017 N. Las Palmas Ave., 

Hollywood, Calif. 
33^1, 73rd St., 

Jackson Heights, L. I., 
N. Y. 


6342 W. 6th St., 

Los Angeles, Calif. 

1935 Fairburn Ave., 

West Los Angeles, Calif. 

112-15, 72nd Rd., 

Forest Hills, N. Y. 
600 S. York St., 

Mechanicsburg, Pa. 
503 Ware Ave., 

East Point, Ga. 
3774 Effingham PI., 
Hollywood, Calif. 
9021 Dicks St., 

West Hollywood, Calif. 
8630 Berri St., 
Montreal, P. Q. 

126 Langdon St., 
Madison, Wis. 
DUKE, D. W. 

Cinema Dept., Univ.of S. Calif. 
Box 326, Los Angeles, Calif. 
DUNN, L. G. 
345 N. Formosa Ave., 

Hollywood, Calif. 
525 Monroe St., 
Corvallis, Ore. 
780 N. Gower St.. 

Hollywood, Calif. 
ELLS, F. C. 
272 Grant St., 

Pasadena, Calif. 

1701 N. Michigan Ave., 

Pasadena, Calif. 
170 Ninth St., 

San Francisco, Calif. 
4537 Placidia Ave., 

North Hollywood, Calif. 


8139 W. 4th St., 

Los Angeles, Calif. 
GUNN, R. L. 

Twentieth Century-Fox Corp., 
10201 W. Pico Blvd., 
Los Angeles, Calif. 

Pathe Labs, of Calif., Inc., 
6823 Santa Monica Blvd., 

Los Angeles, Calif. 

Harrison Projector Co., 
8351 Santa Monica Blvd., 

Hollywood, Calif. 

1544 Midvale Ave., 
West Los Angeles, Calif. 


1514 S. Beverly Drive, 
Los Angeles, Calif. 



[J. S. M. p. E 

HYNE, A. D. 

1429 23rd Ave., 

San Francisco, Calif. 

3662 Dunn Drive, 

Los Angeles, Calif. 

9125 West 25th St., 

Los Angeles, Calif. 
6819 llth Ave., 

Los Angeles, Calif. 
Rialto Bioscope, 

Batavia, Java 

5909 N. Kilbourne St., 

Chicago, 111. 

1554 S. Manhattan PL, 

Los Angeles, Calif. 
LAW, E. A. 

544 N. Alta Vista Blvd., 

Hollywood, Calif. 

c/o W. T. Crespinel, 
4248 Cahuenga Blvd., 

N. Hollywood, Calif. 

2381 Peachtree Rd., 
Atlanta, Ga. 


106 W. 87th St., 

New York, N. Y. 
LUND, S. D. 

3441 Wonder view Dr., 

Hollywood, Calif . 
12040 Otsego St., 

N. Hollywood, Calif. 
V. G. Uberti 7, 
Milano, Italy 

10341 Rossbury PL, 
Los Angeles, Calif. 

McLEAN, F. R. 
Box C, 

Coulterville, 111. 

310 Edgewood Ave., 

West Englewood, N. J. 

526 Arbramar Ave., 

Pacific Palisades, Calif. 
1703 Jefferson St., 
Philadelphia, Pa. 
310 N. 40th St., 

Philadelphia, Pa. 

1216N. KenmoreSt., 
Los Angeles, Calif. 

202-09, 50th Ave., 

Bayside, L. I., N. Y. 

N. Hollywood, Calif. 
RUNK, J. S. 

1511 Clark Ave., 
Burbank, Calif. 
2859 Walnut St., 

Huntington Park, Calif. 
Box 43, 

Shrewsbury, Pa. 

Eastman Kodak Company, 
6706 Santa Monica Blvd. 
Hollywood, Calif. 


1517 Sherbourne Dr., 
Los Angeles, Calif. 

1700 Montana Ave., 

West Los Angeles, Calif. 
40 E. 17th St., 
Brooklyn, N. Y. 



1010 E. 42nd St., c/o Motiograph, Inc., 

Brooklyn, N. Y. 4431 W. Lake St., 

SWAIN, J. J. Chicago, 111. 

1548 N. Orange Grove St., WADE, N. G. 

Hollywood Calif Massachusetts Institute of Tech., 


3535 N. Lakewood St., 

^, . T11 Eastman Kodak Company, 

ChlCag ' IlL 6706 Santa Monica Blvd., 

THORN, T. K. Hollywood, Calif. 

Rialto Bioscope, WEXLER, SY 

Senen ' 901 E. 179th St., 

Batavia, Java New York> N y 


14646 Magnolia Blvd., 119 3. Destnedge Ave., 

Van Nuys, Calif. Kalamazoo, Mich. 


8928 Santa Monica Blvd., 10435 Dunleer Drive, 

Los Angeles, Calif. Los Angeles, Calif. 

In addition, the following applicants have been admitted to the Active grade : 


1207 No. Mansfield Ave., Paramount Pictures, Inc., 

Hollywood, Calif. 5451 Marathon Street, 

Hollywood, Calif. 

I rt r ii AT; 2228 Holly Drive, 

Beverly alls. Calif. Hollywood, Calif. 


c/o Warner Brothers Pictures, Inc., 408 West Cienega St., 
Burbank Calif. San Dimas, Calif. 


Washington Theater Bldg., 
Washington, N. J. 

The following members have been transferred from Associate to Active grade 


United Artists Studio Corp., Consolidated Film Industries, Inc., 

1041 N. Formosa Ave., 959 Seward Street ' 

Hollywood, Calif. Hollywood, Calif. 


SLYFIELD, C. O. Mitchell Camera Corp., 

3216 W. 78th Place, 665 N. Robertson Blvd., 

Los Angeles, Calif. West Hollywood, Calif. 


Prior to January, 1930, the Transactions of the Society were published quar- 
terly. A limited number of these Transactions are still available and will be 
sold at the prices listed below. Those who wish to avail themselves of the op- 
portunity of acquiring these back numbers should do so quickly, as the supply 
will soon be exhausted, especially of the earlier numbers. It will be impossible 
to secure them later on as they will not be reprinted. 


















Beginning with the January, 1930, issue, the JOURNAL of the Society has been 
issued monthly, in two volumes per year, of six issues each. Back numbers of 
all issues are available at the price of $1.00 each, a complete yearly issue totalling 
$12.00. Single copies of the current issue may be obtained for $1.00 each. 
Orders for back numbers of Transactions and JOURNALS should be placed through 
the General Office of the Society and should be accompanied by check or money- 


The following are available from the General Office of the Society, at the prices 
noted. Orders should be accompanied by remittances. 

Aims and Accomplishments. An index of the Transactions from October, 
1916, to December, 1929, containing summaries of all articles, and author and 
classified indexes. One dollar each. 

Journal Index. An index of the JOURNAL from January, 1930, to December, 
1935, containing author and classified indexes. One dollar each. 

Motion Picture Standards. Reprints of the American Standards and Recom- 
mended Practice as published in the March, 1941, issue of the JOURNAL; 50 cents 

Membership Certificates. Engrossed, for framing, containing member's name, 
grade of membership, and date of admission. One dollar each. 

Journal Binders. Black fabrikoid binders, lettered in gold, holding a year's 
issue of the JOURNAL. Two dollars each. Member's name and the volume 
number lettered in gold upon the backbone at an additional charge of fifty cents 

Test- Films. See advertisement in this issue of the JOURNAL. 




Volume XXXVI April, 1941 



A 200-Mil Variable-Area Modulator 


Determination of Microphone Performance 


Scene-Slating Attachment for Motion Picture Cameras 

F. C. GILBERT 355 

A Monochromatic Variable- Density Recording System 


Some Observations on Latent Image Stability of Motion 


Negative Exposure Control D. NORWOOD 389 

Hollywood's Low-Temperature Sound-Stage. .R. VAN SLYKER 403 

Theory Vs. Practice F. H. RICHARDSON 411 

Pioneering in the Talking Picture W. E. THEISEN 415 

Officers and Governors of the Society 445 

Committees of the Society 448 

Current Literature 453 

1941 Spring Convention at Rochester, N.Y., May 5th-8th, Incl. 455 

Society Announcements 459 

Constitution and By-Laws of the Society. 461 





A. C. DOWNES, Chairman 




Subscription to non-members, $8.00 per annum; to members, $5.00 per annum, 
included in their annual membership dues; single copies, $1.00. A discount 
on subscription or single copies of 15 per cent is allowed to accredited agencies. 
Order from the Society of Motion Picture Engineers, Inc., 20th and Northampton 
Sts., Easton, Pa., or Hotel Pennsylvania, New York, N. Y. 

Published monthly at Easton, Pa., by the Society of Motion Picture Engineers. 

Publication Office, 20th & Northampton Sts., Easton, Pa. 

General and Editorial Office, Hotel Pennsylvania, New York, N. Y. 

West Coast Office, Suite 226, Equitable Bldg., Hollywood, Calif. 

Entered as second class matter January 15, 1930, at the Post Office at Easton, 
Pa., under the Act of March 3, 1879. Copyrighted, 1941, by the Society of 
Motion Picture Engineers, Inc. 


** President: EMERY HUSE, 6706 Santa Monica Blvd., Hollywood, Calif. 
** Past-President: E. ALLAN WILLIFORD, 30 East 42nd St., New York, N. Y. 
**Executive Vice-President: HERBERT GRIFFIN, 90 Gold St., New York, N. Y. 

^Engineering V ice-President: D. E. HYNDMAN, 350 Madison Ave., New York. 
** Editorial Vice-President: ARTHUR C. DOWNES, Box 6087, Cleveland, Ohio. 

*Financial Vice-President: A. S. DICKINSON, 28 W. 44th St., New York, N. Y. 
**Convention Vice-President: W. C. KUNZMANN, Box 6087, Cleveland, Ohio. 

* Secretary: PAUL J. LARSEN, 44 Beverly Rd., Summit, N. J. 

^Treasurer: GEORGE FRIEDL, JR., 90 Gold St., New York, N. Y. 


**M. C. BATSEL, Front and Market Sts., Camden, N. J. 

*J. A. DUBRAY, 1801 Larchment Ave., Chicago, 111. 

*JOHN G. FRAYNE, 6601 Romaine St., Hollywood, Calif. 

*A. N. GOLDSMITH, 580 Fifth Ave., New York, N. Y. 

*ARTHUR C. HARDY, Massachusetts Institute of Technology, Cambridge, Mass. 
**L. L. RYDER, 5451 Marathon St., Hollywood, Calif. 

*T. E. SHEA, 195 Broadway, New York, N. Y. 

*R. O. STROCK, 35-11 35th St., Astoria, L. I., N. Y. 

Term expires December 31, 1941. 
**Term expires December 31, 1942. 


Summary. A modulator using a new vibrating-mirror unit has been developed 
for recording double-width variable-area sound-track. The noise-reduction shutter is 
at the slit, making it possible to record, with noise-reduction, class A push-pull track 
comprising two standard bilateral tracks, one of which is located in accordance with the 
dimensional standards for single track. While this has been its principal use to date, 
it is readily adaptable for other types of track. A visual monitor shows operation of the 
noise-reduction shutter and the amplitude of signal modulation in both directions from 
the base-line with a positive indication of peak over-load. An exposure meter is in- 
cluded to serve as a check on lamp current and track balance. The light-source is a 
tungsten filament lamp which will properly expose fine-grain emulsions to "white" 
light or standard emulsions through an ultraviolet filter . 

Equipment has been available and in use for some time for the 
recording of double-width or 200-mil push-pull variable- density 
sound-track. The generally satisfactory results obtained by this 
method have encouraged the development of equipment for making 
the same advantages available for the recording of variable-area 
track. The modulator described herein has been used for recording 
double-width class A push-pull variable-area sound-track although 
it is readily convertible to the recording of other types of sound-track. 

If double-width track is combined with push-pull recording, the 
following advantages are realized. 

(1) The signal-to-noise ratio of a sound-on-film record is increased 3 db. 

(2) Clipping is reduced because faster operation of noise-reduction is permis- 

(5) "Hush-hush" is reduced because, with increased speed of noise-reduction, 
less margin is required. 

(4) Distortion is reduced by the balancing action of push-pull. 

(5) Film processing is less critical for a given degree of distortion. 

Also, if one of the halves of the push-pull track is located in accord- 
ance with single-track standards the double-width push-pull track 

* Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received Decem- 
ber 12, 1940. 

** Electrical Research Products, Inc., Hollywood, Calif. 

$ The Society is not responsible for statements by authors <" 





A. C. DOWNES, Chairman 




Subscription to non-members, $8.00 per annum; to members, $5.00 per annum, 
included in their annual membership dues; single copies, $1.00. A discount 
on subscription or single copies of 15 per cent is allowed to accredited agencies. 
Order from the Society of Motion Picture Engineers, Inc., 20th and Northampton 
Sts., Easton, Pa., or Hotel Pennsylvania, New York, N. Y. 

Published monthly at Easton, Pa., by the Society of Motion Picture Engineers. 

Publication Office, 20th & Northampton Sts., Easton, Pa. 

General and Editorial Office, Hotel Pennsylvania, New York, N. Y. 

West Coast Office, Suite 226, Equitable Bldg., Hollywood, Calif. 

Entered as second class matter January 15, 1930, at the Post Office at Easton, 
Pa., under the Act of March 3, 1879. Copyrighted, 1941, by the Society of 
Motion Picture Engineers, Inc. 


**President: EMERY HUSE, 6706 Santa Monica Blvd., Hollywood, Calif. 
** 'Past-President: E. ALLAN WILLIFORD, 30 East 42nd St., New York, N. Y. 
**Executive Vice-President: HERBERT GRIFFIN, 90 Gold St., New York, N. Y. 

* Engineering Vice-President: D. E. HYNDMAN, 350 Madison Ave., New York. 
**Editorial Vice-P resident: ARTHUR C. DOWNES, Box 6087, Cleveland, Ohio. 

*Financial Vice-President: A. S. DICKINSON, 28 W. 44th St., New York, N. Y. 
** Convention Vice-President: W. C. KUNZMANN, Box 6087, Cleveland, Ohio. 

* Secretary: PAUL J. LARSEN, 44 Beverly Rd., Summit, N. J. 

^Treasurer: GEORGE FRIEDL, JR., 90 Gold St., New York, N. Y. 


**M. C. BATSEL, Front and Market Sts., Camden, N. J. 

*J. A. DUBRAY, 1801 Larchment Ave., Chicago, 111. 

*JOHN G. FRAYNE, 6601 Romaine St., Hollywood, Calif. 

*A. N. GOLDSMITH, 580 Fifth Ave., New York, N. Y. 

*ARTHUR C. HARDY, Massachusetts Institute of Technology, Cambridge, Mass. 
**L. L. RYDER, 5451 Marathon St., Hollywood, Calif. 

*T. E. SHEA, 195 Broadway, New York, N. Y. 

*R. O. STROCK, 35-11 35th St., Astoria, L. I., N. Y. 

*Term expires December 31, 1941. 
**Term expires December 31, 1942. 


Summary. A modulator using a new vibrating-mirror unit has been developed 
for recording double-width variable-area sound-track. The noise-reduction shutter is 
at the slit, making it possible to record, with noise-reduction, class A push-pull track 
comprising two standard bilateral tracks, one of which is located in accordance with the 
dimensional standards for single track. While this has been its principal use to date, 
it is readily adaptable for other types of track. A visual monitor shows operation of the 
noise-reduction shutter and the amplitude of signal modulation in both directions from 
the base-line with a positive indication of peak over-load. An exposure meter is in- 
cluded to serve as a check on lamp current and track balance. The light-source is a 
tungsten filament lamp which will properly expose fine-grain emulsions to "white" 
light or standard emulsions through an ultraviolet filter. 

Equipment has been available and in use for some time for the 
recording of double-width or 200-mil push-pull variable- density 
sound-tracjc. The generally satisfactory results obtained by this 
method have encouraged the development of equipment for making 
the same advantages available for the recording of variable-area 
track. The modulator described herein has been used for recording 
double-width class A push-pull variable-area sound-track although 
it is readily convertible to the recording of other types of sound-track. 

If double-width track is combined with push-pull recording, the 
following advantages are realized. 

(1) The signal-to-noise ratio of a sound-on-film record is increased 3 db. 

(2) Clipping is reduced because faster operation of noise-reduction is permis- 

(5) "Hush-hush" is reduced because, with increased speed of noise-reduction, 
less margin is required. 

(4) Distortion is reduced by the balancing action of push-pull. 

(5) Film processing is less critical for a given degree of distortion. 

Also, if one of the halves of the push-pull track is located in accord- 
ance with single-track standards the double-width push-pull track 

* Presented at the 1940 Fall Meeting at Hollywood, Calif.; received Decem- 
ber 12, 1940. 

** Electrical Research Products, Inc., Hollywood, Calif. 


& The Society is not responsible for statements by authors $ 


R. W. BENFER AND G. T. LORANCE [j. s. M. p. E. 

can be handled on standard reproducing equipment for such purposes 
as cutting and editing, although under these circumstances the opera- 
tion of the noise-reduction system may be audible and distortion may 
be noticeable. 

The dimensions and location of the sound-track recorded by this 
modulator are shown in Fig. 1, the dark areas corresponding to the 
exposed areas on the negative. The track is shown for three condi- 

FIG. 1. 

Dimensions and location of sound- 

tions: (1) unmodulated without noise-reduction; (2) unmodulated 
with noise-reduction ; (3) completely modulated with noise-reduction 
shutter withdrawn by action of the signal. 

A plan of the optical system less the visual monitor is shown sche- 
matically in Fig. 2. 

The light-source is a 10-volt, 7.5-ampere or a 10.5-volt, 7.8-ampere 
curved-coil filament lamp. The condenser lens is a high-aperture 
single-element lens with aspheric surfaces. The combination of con- 
denser and relay lens completely fills the vibrator mirror with an image 
of the filament. The mask, which provides the modulating edges, is 

April, 1941 ] A 200-MlL VARIABLE- AREA MODULATOR 333 

adjacent to the condenser lens and is curved to conform to the curva- 
ture of field characteristics of the relay lens. An image of the mask, 
magnified 1.5 times, is formed by the relay lens at the slit. 

Light to form the image of the mask must be reflected by the 
mirror of the modulating unit, which is positioned so that the axis of 
the reflected beam is at an angle of 30 degrees with the axis of the inci- 
dent beam. To modulate the sound-track the mirror vibrates about a 
horizontal axis, thereby moving the image of the mask triangles up 
and down across the slit. The mirror is the aperture stop of the sys- 
tem. The lamp, condenser lens, and relay lens are designed to supply 
all the light-rays that can be passed by the mirror, collector, slit, and 
objective lens onto the film. 


FIG. 2. Plan of optical system (less visual monitor). 

The slit is accurately engraved in the silvered surface of a thin 
glass "flat," which is cemented to the plane surface of the collector 
lens, the convex side of which faces the mirror. This provides protec- 
tion by glass on both sides of the slit, thus sealing it against dirt. 
The glass surfaces can be cleaned easily with no risk to the slit edges. 
A matte surface is provided by this construction on which the image 
of the mask triangles is made clearly visible. 

Noise-reduction is obtained by intercepting the light emerging from 
the slit with a shutter which is moved into the light-beam in 
accordance with the currents from the noise-reduction circuit. This 
location of the noise-reduction shutter was chosen in order to obtain 
the advantages of a bilateral track in each of the two halves of the 
push-pull track. 



The objective lens, working at a reduction of 8:1, projects onto the 
film a slit image, the nominal dimensions of which are 0.200 inch by 
0.00025 inch. Two appropriately corrected, interchangeable objec- 
tives are available, one for the exposure of fine-grain emulsion to 
"white" light and the other for the exposure of standard emulsions to 
"ultraviolet" light. Both lenses have a working speed on the film 
side of //2.0 for their full circular apertures. Either of the previously 
mentioned lamps, operated at less than rated current and used in this 
system with the appropriate objective lens, will properly expose the 
films mentioned. 

FIG. 3. Vertical section, unfolded, with diagrams of masks, noise-reduction 
shutter, and visual monitor patterns. 

In case a filter, such as the Corning 584, is used, it is mounted near 
the objective lens on the side toward the slit. This location, near the 
lens, minimizes the effects of filter irregularities and dust because of 
its "out-of -focus" position, and it in no way interferes with visual 
observation of any operation or adjustment. 

The light normally incident on the slit can be diverted to an ex- 
posure meter by swinging a mirror into the position indicated. 

Fig. 3 is a vertical section through the optical axis showing an un- 
folded optical schematic as if the mirror were a transparent aperture. 
The visual monitor edges of the mask are imaged by the relay lens at 
an aperture in the plane of the slit above the image of the modulating 
edges. An indicator on the noise-reduction shutter intercepts some 



of this light and is imaged, along with the image of the mask, on the 
visual monitor screen which is located conveniently several inches 
above the optical axis of the recording beam. The light is directed 
and the magnified image is formed by the combination of two prisms 
and a lens as indicated. This image shows the position of the noise- 
reduction shutter and indicates amplitude of signal modulation. 

Illustrated at A (Fig. 3) is the monitor image in "standby" position. 
The edges of the shadow band are defined, respectively, by the peak 
amplitude of each half of the signal wave and travel in opposite di- 

FIG. 4. Modulator, cover removed. 

rections, thus decreasing the shadow width with an increasing ampli- 
tude of signal until, at 100 per cent modulation, the two edges are 
coincident and the shadow band disappears as illustrated at B. At 
higher amplitudes the edges pass each other, thus giving an instan- 
taneous indication of overshooting, which can be evaluated by refer- 
ence to an index on the monitor screen. The pattern for a peak over- 
shoot is shown at C (Fig. 3) . 

Fig. 4 is a photograph of the modulator with the cover removed. 
The lamp, the condenser lens, and the masks comprise a unit assembly 
which includes adjustments for positioning and focusing the lamp 
filament on the vibrator mirror. Lateral 'adjustment of the two 
modulating mask triangles permits registering their images with the 

336 R. W. BENFER AND G. T. LORANCE [j. s. M. p. E. 

noise-reduction shutter at the slit, and rotational adjustment permits 
control of the width of the unmodulated track. The width of the 
monitor mask can be adjusted to agree with the limits for a fully 
modulated track. 

The objective and relay lenses are provided with focusing mounts 
in which the lenses may be moved along the optical axis without rota- 
tion and which permit adjustment and locking without the use of 

FIG. 5. Vibrator unit and cradle assembly. 

A push-button-controlled deflector intercepts the light normally 
incident on the slit and directs it to the sensitive surface of a General 
Electric light-meter. The mask triangles are imaged on this surface 
and a selector is provided so that the illumination available for the 
exposure of each track as well as the total of both tracks can be read 
on the meter. Apertures are provided in each half of the beam to 
adjust the meter for equal deflections when the two track densities 
are identical. Release of the deflector control-button automatically 
restores the light-path to the recording position. 

Fig. 5 is a photograph of the vibrator unit mounted in a cradle 
assembly which provides pivots on the horizontal and vertical axes 
of the mirror. These adjustments are independent and permit regis- 

April, 1941 ] A 200-MiL VARIABLE-AREA MODULATOR 337 

tering the mask images with the noise-reduction shutter to obtain 
symmetrical bias lines. Vibrator units are interchangeable and read- 
ily aligned for operation by reference to index marks on the visual 
monitor screen. Operation of a "quick- tilt" lever rotates the mirror 
about its horizontal axis to expose one track to its full width. The 
vibrator mirror automatically returns to "standby" position when 
the "quick- tilt" lever is released, thus making it convenient to expose 
laboratory test strips whenever desired. 

FIG. 6. Visual monitor, noise-reduction shutter, and slit assembly. 

Fig. 6 is a photograph of the visual monitor, noise-reduction shut- 
ter, and slit assembly. Adjustments are provided that permit align- 
ing the shutter parallel to and in correct register with the slit openings. 
Adjustments for azimuth relative to the film are also provided. The 
visual monitor optical system is a unit assembly supported by the slit 
mounting bracket. Its upper prism can be positioned at 90- degree 
intervals so that the monitor screen can be located in any one of four 
positions. The noise-reduction unit is an electro dynamic structure 
comprising a permanent magnet to supply the field and a moving coil 
for driving the shutter. Springs, one on e.ach end of the magnetic 
structure, are rigidly connected by a rod through the center pole-piece. 
This rod carries the moving coil and is extended upward to drive the 


R. W. BENFER AND G. T. LORANCE [j. s. M. P. E. 

shutter. The shutter is contoured to provide two opaque triangles 
for noise-reduction and the indicator imaged on the visual monitor 
screen. The moving coil of the noise-reduction shutter was designed 

: law - - " ~ ; 

FIG. 7. Complete modulator, covered. 

to operate within the power output limits of present noise-reduction 
circuits and will operate at attack speeds of 8 milliseconds or more as 
determined by the constants of the noise-reduction circuit. 



FIG. 8. Average "U. V." print response ("U. V." negative) for 
constant amplitude of mirror vibration. 

Fig. 7 shows the modulator enclosed in its protective cover, which 
is removable to provide access to the adjustments necessary for in- 
stallation. The controls essential to recording, such as the "quick- 
tilt" lever, the "standby" locking adjustment on the vibrator cradle, 

April, 1941 ] A 200-MiL VARIABLE- AREA MODULATOR 339 

and the controls for the exposure meter are accessible for operation 
without removing the cover. The entire unit rests on a sub-plate 
with a tongue-and-groove screw-thread adjustment for lateral posi- 
tioning of the optical axis in reference to the film. This adjustment 
moves the entire modulator parallel to the film surface and does not 
affect other adjustments. External circuits are connected to a ter- 
minal strip at the rear of the unit. 

The curve of Fig. 8 shows an average response of "ultraviolet" 
prints on EK-1301 for constant amplitude of mirror vibration. Nega- 
tives were on EK-1357 and exposed through a 1-mm Corning 584 
filter and the ultraviolet-corrected objective lens. 

The advantages listed above which have heretofore been restricted 
to variable-density are with this modulator realized in variable-area. 
The laboratory model has been used for recording original sound- 
track for several studio productions. 

Though it is not possible to name all who have aided in this de- 
velopment we wish to acknowledge the contributions of Mr. R. Wolf 
and the cooperation of Mr. Foster of the Bausch & Lomb Optical 



MR. KELLOGG : Can you tell us more about the damping of the shutter and the 
galvanometer, and the manner in which the requisite stiffness is supplied to the 
galvanometer to give it a suitable natural frequency. It looks as if there were a 
tension spring across the window in which the mirror and armature are mounted, 
but I should be glad to have more details. 

Also, I am interested in the reasons for using the large mirror and moving it 
farther from the slit. How does such an arrangement compare with a smaller 
mirror, closer up? I have always felt that the large mirror arrangement takes 
much more input.. For the same amount of light modulation, there is more 
kinetic energy and that bears somewhat upon the power requirements, in which 
we are much interested; although I grant that more power than is at present sup- 
plied is not a great obstacle, if one benefits by it. 

MR. BENFER: The armature is a torsional structure formed from a single 
piece of Permendur. It is separated from the pole-pieces by a tungsten wire on 
which it pivots. The mirror is cemented directly to the armature. Damping 
is obtained magnetically within the modulator itself and is controlled by selec- 
tion in design, magnetic materials, and flux densities. 

MR. LORANCE : Any device of this type has a response characteristic that is a 
function of the impedance out of which it is driven. The coupling network 
smooths out the impedance irregularities and provides equalization for the re- 
sponse characteristic. The impedance equalizer section is wasteful of power be- 
cause resistance is put in series with the vibrator coiHn order to simplify the net- 
work. About +28 to +29 db relative to 6 milliwatts is required at the network 
input for full modulation. 


Damping is obtained chiefly through the design constants of the magnetic 
structure and is quite good as indicated by the relatively few free vibrations, or 
transients, accompanying a keyed signal. 

The size of the mirror is related to the whole modulator by the optical working 
speed desired and by the magnifications or reductions involved. Practical con- 
siderations also have something to do with it. In the preliminary layouts of the 
modulator it appeared wise to use a rather large slit so that a rather large mask 
image would be available for observation in the plane of the slit, and so that com- 
ponents would be at distances that would not interfere with adjustment or ob- 

MR. KREUZER : Has the exposure meter been corrected for temperature ? 

MR. LORANCE : No, it has not. It is included chiefly as a guide and as a check 
on lamp current. It is not a substitute for a lamp current meter. 

MR. OFFENHAUSER: About what is the density of the slit in the opaque re- 

MR. LORANCE: We have never tried to check it; but there has been no trace 
of useless light coming through. It is thick enough to be "opaque." 


Summary. Methods of determining the performance characteristics of micro- 
phones by acoustic measurements are described. Work Factors involving the accuracy 
of the methods are discussed. The correlation between a microphone's performance 
as determined by acoustic measurement and by listening tests is reported. Applica- 
tion of both types of test to a studio type of cardioid microphone is given as an example. 

The performance characteristics of microphones are of considerable 
interest to studios since the ultimate sound quality of the recording 
system is a function of any limitations that the microphone may im- 
pose. Factors influencing the choice of a microphone for sound re- 
cording will not be considered here, since they have previously been 
described. 1 

Generally, two methods of judging a microphone's performance are 
available: listening tests and acoustic measurements. 9 Properly 
conducted listening tests can give considerable information regarding 
the microphone under actual conditions of use. Usually, comparisons 
are made between microphones of a known and used type, and others 
whose performance is unknown. The basic philosophy for such a 
test should be to determine which microphone gives the most faithful 
reproduction of the original speech or music. However, the perform- 
ance of the new microphone is often contrasted with the results given 
by the old type, without reference to the original source of sound. 
Such a procedure is likely to be misleading. It would appear more 
straightforward to select the best microphone on a basis of comparison 
with the original sound-source, and then modify its performance char- 
acteristic, if necessary, acoustically or electrically, to secure the de- 
sired characteristic. This latter procedure may be justified at times 
due to other equipment limitations, or to a desire to achieve some 
particular type of sound quality for dramatic effect. Such listening 
tests for judging the microphone's performance are affected by stage 

* Presented at the 1940 Fall Meeting at Hollywood, Calif.; received October 
15, 1940. 

** Electrical Research Products, Inc., Hollywood, Calif. 


$ The Society is not responsible for statements by authors $ 

342 F. L. HOPPER AND F. F. ROMANOW [j. s. M. p. E. 

acoustics as well as the characteristics of the associated transmission 
equipment, including recording or reproducing systems and the 
monitoring loud speaker. Inasmuch as this type of test represents 
the way in which a microphone is used, it is probably an effective way 
in which to judge actual performance. 

Contrasted to this type of test is the determination of the micro- 
phone's response characteristics by means of acoustic measurements. 
Here the aim is, so far as possible, to remove all extraneous acoustical 
effects and characteristics of equipment. The resulting data show 
the inherent response of the microphone in a specific type of sound 
field. This type of test is particularly useful to the designer of micro- 
phones as a yardstick to judge his achievements. 

If similar test equipments of comparable performance are available 
to the designer, manufacturer, and user, a common basis exists for a 
discussion of microphone problems. Similar test equipments, facili- 
ties, and technics are employed by Bell Telephone Laboratories and 
the Western Electric Company, and recently a comparable test set-up 
has been established at Electrical Research Products' West Coast 
laboratories. This system is available for the following purposes : 

(1) To serve as a reference. As such, it may be used to correlate and unify the 
results of various microphone test methods now employed by studios. 

(2} To determine the effects of acoustic baffles and wind screens, proposed or 
employed by studios, upon the normal characteristics of a given microphone. 

(3} To measure and compare the response characteristics of microphones of a 
given type and to check the performance of those suspected of trouble. 

(4} To correlate measured response with comments on a microphone's per- 
formance as judged from listening tests, leading perhaps to a better understanding 
of future design requirements. 

The response of a microphone is defined as the ratio of the open- 
circuit voltage generated at the terminals of the instrument to the 
acoustic pressure actuating the device. Two methods of measure- 
ment are generally available. In the first, or "pressure" calibration, 
the pressure is uniform over the diaphragm and is measured at the 
diaphragm. In the second method, or "field" calibration, the mi- 
crophone is introduced into the field of a plane progressive sound- 
wave. The presence of the microphone distorts the sound field, and 
the pressure on the diaphragm of the microphone is different from 
that existing before the microphone was placed into the field, due to 
diffraction and resonance effects around the microphone. In the 
"field" calibration of the microphone the undisturbed sound pressure, 


that is, the pressure before the microphone is placed into the sound 
field, is taken as the input pressure. The field response of a micro- 
phone is, therefore, the open-circuit voltage divided by the undis- 
turbed sound pressure at the microphone position. 

An example of a "pressure" calibration is illustrated by the thermo- 
phone method 8 of calibrating condenser transmitters. It is primarily 
useful in establishing the characteristics of a condenser transmitter 
that may subsequently be used in making field calibrations of other 
microphones. In this method, the diaphragm of the transmitter 
constitutes one wall of a small cylindrical enclosure. Inside this 
chamber are mounted two strips of gold foil through which alternating 
and direct currents are passed. The cyclic changes in the alternating 
current produce corresponding changes in the temperature of the foil, 
and therefore an alternating pressure distributed uniformly in the 
chamber is created to actuate the transmitter diaphragm. In order 
to prevent resonances from occurring in the chamber at frequencies 
at which the pressure calibration is made, it is filled with hydrogen 
gas, which displaces such irregularities upward in the frequency range. 

Since the field calibration method more nearly corresponds to the 
performance of a microphone under conditions of use, it is of more 
general interest. A field calibration of a commercial type microphone 
is generally made by a comparison method. A reference micro- 
phone whose field characteristic has been established by basic meth- 
ods is introduced into a sound field. Its output may be measured and 
drawn automatically by means of a sound-level recorder. The un- 
known microphone is then similarly measured in the same sound field 
and the difference in response is applied to the field calibration of the 
known microphone, thus determining the field calibration of the un- 
known microphone. 

A number of precautions must be taken if accuracy is to be achieved. 
A fundamental requirement is the provision of a stable calibrated 
reference microphone. The miniature condenser transmitter 640- A 
probably represents the best obtainable microphone for this purpose. 
An exact field calibration of the miniature transmitter is necessary. 
This is done by using a field of a plane progressive sound-wave and 
measuring its pressure by means of a Rayleigh disk. 2 Briefly, this 
method depends upon the fact that a light disk suspended in a sound 
field tends to orient itself so that its plane is perpendicular to the 
particle displacement in the wave. Data obtained from such a de- 
vice permit a mathematical calculation of the sound-field pressure to 


F. L. HOPPER AND F. F. ROMANOW [j. s. M. P. E. 

be made. A measurement of the output of the condenser transmitter 
when placed in this known sound field then gives the data for its field 
response. This calibration, when compared to a pressure calibration 
of the same transmitter by the thermophone method, shows a marked 
increase in the output at the upper frequency range. The increase is 
caused by two factors: (1) the diffraction caused by the microphone, 
and (2) the resonance of the small cavity in front of the diaphragm. 
The diffraction effect of a circular disk has been given in a recent 
paper 3 and is shown in Fig. 1. To this must be added the resonance 
effect due to the small cavity in front of the diaphragm which can be 
calculated. 4 The sum of these two corrections is added to the ther- 


FIG. 1. Diffraction effect of a circular disk. 

mophone calibration of the 640-A transmitter to give the theoretical 
field response. Fig. 1 compares this response with the Rayleigh disk 
calibration, and it is found that good agreement exists between the 
two methods of arriving at the field calibration. The difference be- 
tween field and thermophone response, sometimes referred to as a 
field correction is a constant for this type of microphone, since the 
effect is a function of the physical shape and size alone and does not 
depend upon the electrical properties of the transmitter. Since a 
Rayleigh disk calibration is laborious and time consuming, and sub- 
ject to disturbances by extremely light air currents in the testing 
room, it is more practicable to calibrate the individual miniature 
transmitters with the thermophone and to apply the field correction 
previously determined for this type. 


In using miniature condenser transmitters an electrical correction 
is also required. A condenser transmitter with small active capacity 
experiences a coupling loss when connected to an input circuit of a 
vacuum tube. To measure this effect, a voltage is placed in series 
with the microphone and the output voltage is observed. Then the 
transmitter is disconnected, and the same voltage as before is applied 
to the input terminals of the system. The output voltage is observed 
again. The difference between the two output voltages in db is the 

FIG. 2. Sound-treated room of the Bell Telephone Laboratories. 

coupling loss, and is a measure of the voltage generated at the grid of 
the tube by the transmitter as compared to its open-circuit voltage. 
This correction is necessary since in calibrating microphones it is cus- 
tomary to compare the voltage which an unknown microphone gener- 
ates at the grid with that oi the reference microphone also at the grid. 
If the unknown microphone is of sufficiently high impedance com- 
pared to the input impedance of the test amplifier, a similar procedure 
is necessary to obtain its open-circuit voltage, which is the quantity 
needed for stating the response. Periodic checks of the condenser 
transmitter by the thermophone method, and an electrical check of 

346 F. L. HOPPER AND F. F. ROMANOW [J. S. M. p. E. 

the amplifier circuit into which it operates insures detection of any 
change in the standard. 

It is obvious that if the sound field into which the microphone is 
placed is non-uniform, errors will be introduced due to an inability to 
separate distortions produced in the sound field by the microphone 
from those inherent in the field itself. Disturbances in the sound 
field around the microphone that are not due to the microphone itself 

FIG. 3. Outdoor set-up 

are those due to reflections from surrounding surfaces. Most of these 
effects may be minimized by utilizing sound sources that radiate 
principally in one direction, thus limiting reflections to one principal 
surface, by utilizing a room so constructed as to be nearly completely 
absorbing, 5 or by working outdoors. When weather and noise con- 
ditions are severe, use may be made of the sound-treated room. Such 
a room in use at the Bell Telephone Laboratories is shown in Fig. 2. 
However, favorable outdoor conditions in Hollywood led to a decision 
to establish a measuring set-up outdoors; as a consequence, the loud 

April, 1941] 



speakers were installed in the roof of the ERPI Laboratory, pointing 
vertically upward, with a boom structure provided to support the 
microphone above the loud speakers. A wind screen, cheesecloth on 
frames, surrounds the structure and permits measurements to be 
made under varying conditions of wind. The wind screen is suffi- 
ciently remote and transparent acoustically to cause a negligible ef- 
fect upon the sound-field characteristic. This set-up is such that it 
permits calibrations to be made of microphones at varying degrees of 
incidence. With highly directional microphones, the response may 
decrease some 20 to 30 db for certain angles of incidence. The ar- 
rangement mentioned, shown in Fig. 3, permits this to be done. 





Due TO 



FIG. 4. 

Correction for ribbon and cardioid types of 

Due to the method of comparing two microphones, the sending and 
receiving links of the measuring system are common to both sets of 
measurements and in general may be disregarded. Under some cir- 
cumstances, when deriving the response characteristic of one type of 
microphone from that of an entirely different type, certain correc- 
tions are necessary. Among these are : 

(i) Corrections due to a difference in performance when microphones of differ- 
ent types are operated into an input transformer in the measuring system. This 
might be due to a difference in impedance characteristic as a function of frequency 
with which the input transformer is faced; or, in the case of microphones employ- 
ing output transformers, some reaction between the latter and the input trans- 
former of the measuring system. Since it is customary to describe a microphone- 
response characteristic in terms of open-circuit voltage, it should either be meas- 
ured with the microphone connected to the grid of a vacuum tube, or, when oper- 



ating into a transformer, the effects of the transformer should be determined and 
removed from the measurement. Such corrections are usually required in the 
case of cardioid and ribbon microphones. 

In general, when comparing the output obtainable from two electrodynamic 
microphones, it is necessary to consider their impedances. Suppose two micro- 
phones have the same field response but one has a 10-ohm impedance and the 
other a 1000-ohm impedance. We might connect an ideal transformer to the 
terminals of the low-impedance microphone to bring it up to the 1000-ohm impe- 
dance level. This would increase the voltage output of the 10-ohm instrument by 
20 db. Since the open-circuit voltages were equal without the transformer, the 

FIG. 5. Equipment in room below the loud speakers. 

output of the 10-ohm microphone can be increased 20 db. It is apparent from 
this example that a statement of the field response of a microphone is incomplete 
as an indication of the obtainable voltage output unless its impedance is also 

(2) When measuring at low frequencies with microphones operating wholly or 
partially upon a pressure-gradient 9 basis, a correction must be made due to the 
increase in amplitude and phase-shift of the particle velocity relative to pressure. 
Were the sound radiator capable of supplying a plane wave, this would not be 
necessary; but with the usual types of loud speaker and the close distances at 
which measurements are made, the wave-form is approximately spherical and a 
correction is required. The nature and magnitude of this correction for ribbon- 


and cardioid-type microphones are shown in Fig. 4. For example, if a ribbon 
microphone is 6 inches away from a spherical sound-source its output is increased 
by 11.5 db at 100 cycles over that with a plane-wave source. For the same in- 
crease a cardioid microphone would have to be nearer the sound-source, namely, 3 
inches away. 

Actual measuring equipments, while varying in detail, generally 
operate in much the same manner. The following brief description 
covers the ERPI test equipment utilized in Hollywood. As has been 
mentioned, the loud speakers and microphones are located on the 
roof of the Laboratory with measuring equipment located in a room 
below. The loud speaker system consists of a low-frequency dy- 
namic-type unit utilized for frequencies from 50 to 500 cycles. A 
moving-coil type of receiver unit is coupled to an acoustic impedance 
element (tubular microphone) and is utilized as a radiator for fre- 
quencies from 500 to 10,000 cycles. In this instance the tubular 
structure is employed as a directional radiator of sound rather than 
in its usual role as a sound collector. Its advantages are two-fold: 
it radiates directionally, thereby reducing reflections from nearby 
objects, and it provides a sound field of very uniform character, in one 
direction. Generally, the pressure at any point in the sound field 
in the vicinity of the microphone test position does not deviate more 
than 1 db from an average value. 

The remainder of the equipment is located in the room beneath the 
loud speakers and the principal components are shown in Fig. 5. A 
beat-frequency oscillator equipped with a motor-driven sweep sup- 
plies frequencies to a power amplifier, volume indicator, and the loud 
speakers. The standard or unknown microphone's output is applied 
to a two-stage amplifier. This amplifier is designed to operate di- 
rectly from a microphone, or from a microphone and associated 
amplifier, as in the case of the condenser transmitter. The amplifier 
operates directly a volume indicator having a logarithmic response 
characteristic, i. e., its deflections are linear in db. The output cur- 
rent of the volume indicator supplies a d-c recording type of milli- 
ammeter. Combined with the sweep features of the oscillator is a 
contacting device which marks the position of a number of frequen- 
cies upon the recorder chart through the use of an auxiliary pen lo- 
cated in the margin of the chart. The system is automatic in its 
operation. The response characteristics of the loud speakers are suf- 
ficiently uniform so that an inspection of the chart gives a rough ap- 
proximation of the microphone response characteristic. Sufficient 


F. L. HOPPER AND F. F. ROMANOW [J. s. M. P. E . 

sound energy output can be secured from the loud speakers so that 
the signal-to-random-noise ratio at the microphone does not limit 
measurements of 90- or 180-degree response in pressure gradient of 
cardioid microphones. 

Utilizing a calibrated, miniature-condenser transmitter and the de- 
scribed system, response characteristics have been derived for a 
number of different types of microphones that had been previously 
calibrated by similar methods at the Bell Telephone Laboratories. 
The results of these two calibrations on the average agree within 1 db, 
with occasional maximum disagreements of the order of 2 db. This 
agreement is considered excellent for this type of measurement. 





FIG. 6. Normal incidence response characteristic for the 
cardioid condition. 

Having discussed methods of determining a microphone's character- 
istic by acoustic measurement, it is of interest to relate how the data 
correlate with those obtained from listening tests utilizing speech and 
music as a sound-source. The differences in response characteristics 
of two microphones having different relative amounts of low or high 
frequencies are easily detected by ear. These differences agree with 
the results of acoustic measurement. However, in many instances 
two microphones of the same type may have nearly, or the same, pro- 
portion of low to high frequencies, so that it might be expected that 
they would sound alike. It is common experience that such micro- 
phones may still exhibit noticeable difference in sound quality. The 
difference is often described in terms of smoothness, naturalness, or 
definition. It is apparently a function of the number and amplitude 

April, 1941] 



of the sharp irregularities that occur in most microphone response 
characteristics. A reduction in either of these factors results in an im- 
provement in quality that the ear can appreciate. As a check, lis- 
tening tests have been conducted with a number of microphones of a 
given type, the microphones being rated as to order of preference. 
Acoustic measurements were then made and it was found that the 
best instruments were those that had the smallest irregularities in 
their response characteristic. Thus acoustic measurements become a 

FIG. 7. Two views of the microphone. 

more valuable tool in evaluating performance of microphones under 
design before they reach the field. 

Such factors have been given consideration in the design of a car- 
dioid type of microphone designed primarily for use in motion picture 
sound recording. The cardioid characteristic is obtained by combin- 
ing in series the outputs of a pressure-type element and pressure- 
gradient element. 6 While the cardioid characteristic is obtained 
when equal voltages are supplied by both elements, other useful 
directional patterns may be secured by varying the proportionality 
of voltages taken from the two units. 7 This variable directional pat- 
tern is useful in three-walled or enclosed sets. A certain amount of 

352 F. L. HOPPER AND F. F. ROMANOW [j. S. M. P. E. 

reflected sound may be excluded by the choice of one of the various 
patterns, with a definite improvement in quality. This is particu- 
larly noticeable in the bass register. In this way the new microphone 
accomplishes an effect acoustically that can not be readily achieved 
by electrical equalization in the recording system. 

The normal incidence response characteristic for the cardioid con- 
dition is shown by the upper curve of Fig. 6. For the purpose of a 
listening test, a microphone having such a response was compared 
with a laboratory model that had a somewhat abnormal characteristic, 
as indicated by the lower curve of Fig. 6. It was found that the ear 
could recognize a quality improvement corresponding to the reduc- 
tion in irregularity of response. The shape of the microphone hous- 
ing was designed with this in mind. Its symmetry of shape and the 
silk screen cemented to the housing aided in minimizing wind noise. 
The microphone is illustrated in Fig. 7. The resilient mount- 
ing of the microphone unit within the housing makes the instrument 
less susceptible to mechanical movement when it is used on a micro- 
phone boom or fishpole. 


1 HOPPER, F. L.: "Characteristics of Modern Microphones for Sound Record- 
ing," /. Soc. Mot. Pict. Eng., XXXIII (Sept., 1939), p. 278. 

2 SIVIAN, L. J.: "Rayleigh Disc Method for Measuring Sound Intensities," 
The London, Edinburgh, and Dublin Philosophical Magazine and J. Science, 7th 
Series, V (March, 1928), p. 615. 

3 MULLER, G. G., BLACK, R., AND DAVIS, T. E.: "Diffraction Produced by Ob- 
stacles or Plates," J. Acoust. Soc. Amer., X (July, 1938), p. 6. 

4 HARRISON, H. C., AND FLANDERS, P. B.: "An Efficient Miniature Condenser- 
Microphone System," Bell Syst. Tech. J., XI (July, 1932), p. 451. 

6 BEDELL, E. H.: "Some Data on a Room Designed for Free Field Measure- 
ments," /. Acoust. Soc. Amer., VIII (October, 1936), p. 118. 

6 MARSHALL, R. N., AND HARRY, W. R. : "A Cardioid Directional Microphone," 
J. Soc. Mot. Pict. Eng., XXXIII (Sept., 1939), p. 254. 

7 MARSHALL, R. N., AND HARRY, W. R.: "The New Six-Way Cardioid Direc- 
tional Microphone," Pickups (Western Electric Co.} (May, 1940). 

8 SIVIAN, E. J.: "Absolute Calibration of Condenser Transmitters," Bell Sys- 
tem Tech. J., X (Jan., 1931), p. 96. 

9 "American Recommended Practice for the Calibration of Microphones,' 
Amer. Stand. Assoc. (1938). 


MR. DAILY: Has a correlation been worked out yet between the measured rate 
of change of the response characteristic of a microphone in the region of irregulari- 
ties and the accompanying quality deterioration that takes place? 


MR. HOPPER: I believe not. We have recognized only recently that these ir- 
regularities can be appreciated by the ear. At the present time I do not believe 
we are in a position to describe accurately what sort of irregularity produces a 
given effect. 

MR. LINDSAY: Is the sound-source always warbled in this test? 

MR. HOPPER: No. Measurements were made using a warble only, a sweep 
only, and a combination of both. We found, due to the fairly uniform sound field 
we had, and the absence of surfaces surrounding the sound field, that the sweep 
was quite adequate. 

MR. TASKER: Will you kindly explain in detail the set-up shown in Fig. 3? 

MR. HOPPER: Fig. 3 shows the pipe structure. The microphone is supported 
by the boom arm at a fixed distance from the longest pipe. The sound field at this 
point was explored with the condenser transmitter for uniformity, and found to 
be satisfactory. In high-frequency response measurements the plane of the mi- 
crophone diaphragm is placed at a specified distance from the end of the longest 
pipe. For measurements at various angles of incidence the microphone is rotated 
about a horizontal axis, keeping the center of the diaphragm above the longest 
pipe. For measurements at low frequencies the microphone is centered over the 
low-frequency loud speaker by rotating the boom arm. The pipe structure is 
acoustically coupled to an electrodynamic type of driving unit. 

MEMBER : The curves you showed were for horizontal response ; what about the 
response in a vertical plane? 

MR. HOPPER: They are almost identical. If the cardioid pattern is rotated so 
as to describe a complete solid, the characteristics in the horizontal and verti- 
cal planes will be nearly the same. 

MEMBER: Would there be any difference in response if the microphone were 
suspended so that the unit would hang vertically? 

MR. HOPPER: The position of the microphone is probably limited by the way 
it is used in production. Microphones are generally hung from a boom or fish- 
pole, and are overhead to get them out of the camera line. That necessitates point- 
ing them down at the actors. There is not much occasion to use them in other 

MR. TASKER: The question may have to do with a phenomenon that has been 
commented on from time to time. It has been said that a velocity-type micro- 
phone, if tilted in one direction or another while following the actors, may become 
"bassy." That may not be true, but it has been claimed. 

MR. HOPPER: The ribbon element in this microphone is constructed in such a 
way that, while it is rather rigid, it may still move freely in the magnetic field 
under the influence of an acoustic wave. It is our belief that the response charac- 
teristic of the ribbon is independent of its operating position. 

MR. RYDER: Have any measurements been made to determine the effect, upon 
the measurements, of two frequencies imposed upon the microphone at the same 

MR. HOPPER: Not in that particular way, although we inadvertently made 
measurements with sufficient input to the loud speaker to cause overloading. 
There were many harmonics in the output, but the measured response of the mi- 
crophone was the same. 

MR. RYDER; Has the system been used for calibrating microphones? Is there 


any difference between measurements made outside in summer or winter, or under 
low or high humidity? 

MR. HOPPER: This much has been noted: the pipe structure that is used as a 
high-frequency sound-source is somewhat subject to frequency characteristic 
changes with humidity. In other words, the response of the device as a loud 
speaker does vary somewhat with temperature and humidity. Since the cali- 
brated condenser retains its calibration, it is possible by that means to determine 
whether the speaker system changes. 

The only time such an effect is of consequence is when the speaker system 
changes while the measurements are being made. If the system remains stable 
for the two comparison measurements, that is all that is required. If we make 
measurements over the course of a day, check runs are made with our calibrated 
microphone at sufficient intervals to assure ourselves that the loud-speaker system 
has not changed, or, if it has, to enable a correction to be applied. 

MR. SKINNER: How long does it take to make a measurement of a unit? 

MR. HOPPER: About fifteen minutes to set the equipment up and about a 
minute and a half to make one run. Two runs are required, one on the calibrated 
microphone and one on the unknown microphone. It is done with a continuous 
sweep, and it is recorded by a continuous chart recorder, so it is completely auto- 
matic in its operation. 

MR. SKINNER : Could you not speed it up? 

MR. HOPPER : If the sound field is shifted too rapidly some of the irregularities 
in the response of the microphone will not appear on the chart. There is a definite 
time interval below which one should not go in order to get a true picture of what 
the microphone is doing. 

DR. DAILY: The charting-type meters used in the measurement of response 
characteristics have definite time-deflection characteristics, approaching one 
second for one type in common use. Therefore, if rapid variations in amplitude 
occur, due to irregularities in response of the microphone under test, the needle of 
the meter will not truly follow the change, resulting in an inaccurate charting of 
the characteristic. 

MR. MUEHLER: What is the amount of distortion in this or any other type of 

MR. HOPPER: We have not been successful in making such measurements. 
Usually the harmonic output of the loud-speaker system is a "bottleneck," and 
as a result the measurements do not mean anything. 

Once we tried measurements supplying square waves to the loud-speaker system 
but that was unsuccessful. All the harmonics of the square wave must be repro- 
duced in the proper phase to get the square waves back. Since the loud speaker 
will not do that the resultant patterns did not mean anything. I doubt very 
much whether the harmonic output of the microphone is a very appreciable factor. 

MR. LINDSAY : Can you tell us something about the frequency range of these 
later- type microphones? 

MR. HOPPER: The cardioid microphone was designed to be usable up to 15,000 
cycles with moderate equalization. The response curve is quite uniform up to 
about 8000 or 9000 cycles, requiring about 6 db of equalization to flatten it out at 
15,000 cps. It was intended to be used with the stereophonic equipment, where 
we strive for a wider frequency band. 



Summary. The anticipated reduction of film markets attendant upon distur- 
bances in Europe caused many studios to reexamine production routines and prac- 
tices with a view to reducing costs without impairing quality. 

A routine, in widespread use in much the same manner, which gave promise of 
cost saving was that of marking "takes," at time of photographing, for ready identi- 
fication through subsequent stages of picture production. The process of so marking 
film is referred to, within the studios generally, as "slating." 

Analysis of the shortcomings of the slating method employed by our studio led to 
the development of a slating attachment mounted upon the camera blimp or iris 
rods and operated by the assistant earner man. 

The design requirements formulated and the manner and degree of compliance 
embodied in the device now in production use are described. 

The anticipated reduction of film markets attendant upon the dis- 
turbances in Europe has caused many studios to reexamine produc- 
tion routines and practices with a view to reducing waste. 

A routine in widespread use that gave promise of cost saving was 
that of marking " takes," at time of photographing, for ready identi- 
fication through subsequent stages of picture production. The proc- 
ess of so marking film is referred to within the studios generally as 

The scene-slating method in use heretofore at Paramount has re- 
quired that the assistant cameraman carry out onto the set a slate- 
board some 14 inches high by 12 inches wide bearing the identifying 
legend. In general, the scene and take numbers consisted of printed 
inserts slid into holders provided on the slate. Such other informa- 
tion as might be desired was written or printed upon insert strips for 
which, also, holders were provided on the slate. It was the practice 
to run the camera "wild" from rest to photograph the slate between 
the end of one take and the beginning of the next. The camera was 

* Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received October 
31, 1940. 

** Paramount Pictures, Inc., Hollywood, Calif. 


$ The Society is not responsible for statements by authors $ 

356 F. C. GILBERT [j. s. M. P. E. 

then stopped, restored to the interlock system, interlocked, and the 
start-mark fogged onto the film, whereupon the camera was ready to 
be restarted upon call. The camera, recorder, and, where involved, 
the background projector, start in interlock. Through the auto- 
matic operation of a douser in the recorder, sound-track exposure be- 
gins at the instant the interlocked system starts turning over, thereby 
providing the necessary sound-track synchronizing information. 

The practice of scene-slating between takes rather than upon 
starting in interlock was adopted to eliminate the distraction caused 
by the assistant's running out onto the set and holding up the slate, 
often literally under the noses of the actors. However, the "between 
takes" slating practice is subject to three serious objections: (1) it 
consumes film not otherwise required; (2) where a retake follows im- 
mediately upon the conclusion of the preceding take there is a loss in 
production time wholly assignable to slating; and (3) frequently 
there occurs both time loss and waste of footage over the period re- 
quired for the assistant to move out of the field of the camera. 

The scene-slating procedure, when the attachment to be described is 
used, is as follows: The camera stands "ready," having been inter- 
locked and the start-mark fogged. Upon the order to "turn-over" 
the assistant cameraman presses a lever raising an auxiliary optical 
system to a position on the camera's optical axis and immediately in 
front of the camera or blimp, thereby exposing the film to the slate 
while the motor system is coming up to speed. At the instant the 
signal is received, indicating that the motor system is up to speed, 
the assistant cameraman removes his finger from the slater-operating 
lever, permitting the arm carrying the auxiliary optical system to drop 
back into the slater housing whereupon "action" can commence. 

Assuming that the new procedure, for 50,000 takes per year, saves 
only the cost of 3 1 /* feet of developed negative per slate, together with 
the cost of developed prints for the slates actually printed and de- 
veloped, the new system gives promise of a potential annual reduction 
in film expense of $8000. This figure does not attempt to evaluate 
production time saving or time saving accruing elsewhere throughout 
the various operating phases as a result of the improvement in the 
legibility of the average slate. 

At the outset of the design of the attachment the following require- 
ments were established as necessary or desirable : 

(1) That the device be adaptable to the following types of camera equipment: 
Mitchell JVC Camera 


in blimp, 

on open tripod, 

boom mounted. 
Mitchell Standard Camera 

in booth, 

on open tripod. 
Bell & Howell Camera 

(2) That the slating attachments be interchangeable. 

(5) That the device provide legible complete slates for camera lens focus dis- 
tance settings from 6 feet to infinity when used with lenses of the focal lengths 
commonly employed. 

(4) That a range of illumination adjustment be provided to accommodate 
stop settings between //2.3 and//22. 

(5) That the scene and take numbers as well as other data subject to change 
on each take be determinable at any time by the assistant from inspection of the 
exterior of the device. 

FIG. 1. Slating attachment and battery box. 

(6) That the device be usable on location as well as throughout the studio. 

(7) That the device fall of its own accord out of the range of the camera upon 
completion of the slating operation. 

(8) That weight and bulk be minimized. 

Fig. 1 shows the battery box and the slating attachment in its 
closed position. The mounting brackets, which are attachments 
permanently associated with each of the blimps and cameras of the 
several types, are not shown. Any of the sixteen slating attachments 
manufactured may be used on any of the twenty-nine blimps, booths, 
or cameras of the types mentioned above and now equipped with the 
requisite standardized mounting facilities. The lack of uniformity 
between cameras or blimps of a given type has been accommodated 
in the fabrication of the brackets referred to. 

Battery power supply to operate the lamps used to illuminate the 
slate face was selected in preference to power taken from the motor 
power supply circuit because of the complications arising from the 



tj. S. M. p. E. 

fact that the motor power supply might be 220-volt a-c, or 32-volt, 
or 135-volt d-c, depending upon the particular camera set-up in use. 
Fig. 2 shows schematically the optical layout of the device. The 
face of the slate-box is illuminated by four bulbs located two on each 
side of the slate. Proceeding from the slate face in the direction in- 
dicated by the arrow the image of the slate is reflected up by a second - 
surface mirror, through the first of two focus-correcting lenses onto a 
first-surface mirror located on the optical axis of the camera; thence 
through the second focus-correcting lens and throught the camera 
lens, imaging the slate at the plane of the film. 


FIG. 2. Schematic of optical system. 

A second-surface mirror was used adjacent to the slate because of 
the impracticability, due to its location, of enclosing the mirror in a 
protective housing. The first-surface mirror and the two focus- 
correcting lenses are combined in a dustproof sub-assembly. 

The focus-correcting lens system was calculated to provide the 
best focus of the slate when the camera lens in use is focused upon a 
plane 8 feet 6 inches from the camera. Due to the limited depth of 
field of any lens system this distance, necessarily a compromise, was 
chosen with a view to securing legible slates over the range covering 
the largest possible number of normal shots. Because of the low 
curvature of the focus-correcting lenses, it was not necessary to cor- 

April, 1941] 



rect them chromatically. The curves used were chosen to minimize 
spherical aberration. 

A further necessary consideration was the size of the slate-box face. 
Since the focal distances were established rather closely by mechanical 
considerations, it was necessary to limit the slate face to a size such 
that for the longest focal length lens normally used the image pro- 
duced at the film plane would not be so large as to result in cutting 
off essential printed matter. At the same time it was desirable to 
keep the slate face as large as possible, to the end that with the short- 
est focal length lens the best image legibility might be maintained. 

FIG. 3. Slates obtained for indicated photographic conditions. 

It is our belief that it is more important to avoid the possibility 
that even an occasional scene may be photographed out of focus 
through failure of the assistant properly to reset the camera lens than 
it is to insure, by separate focusing for the slate, that all slates be in 
more perfect focus. Our operating practice as regards the slate is 
based on that premise but may be changed if desired. 

Fig. 3 shows that without refocusing the camera lens good slate 
legibility is maintained for the 35-mm lens for focus adjustments of 
the lens between 5.5 feet and 40 feet when operating at //2.3. The 
legend on the white inserts describes the conditions as to stop setting, 
and distance for which the focus of the lens was set. 



[J. S. M. p. E. 

For camera lenses of focal lengths longer than 35 mm when operat- 
ing at //2.3, the out-of-focus condition of the slate image impairs 
legibility for camera lens focus distances less than 5 x /2 feet and greater 
than 35 feet. 

For all focal length lenses between and including 24 mm and 100 
mm slate size is satisfactory. Legibility as far as focus is concerned 

FIG. 4. (Upper} Front view of attachment showing operating arm blocked 

in its "raised" position and slate-box removed. 

FIG. 5. (Lower] Attachment with slate-box inserted, arm "raised" and one 
of the lamp housings swung open and lamps dismouned. 

is satisfactory for lens focus distances from 4 feet to infinity provided 
the stop is//4.5 or above. 

The seriousness of the limitation on the longer focus distances is 
alleviated in practice by the fact that most of the takes involving ad- 
justment of the camera lens to distances of 35 feet or more are out- 
of-doors where the use of stop settings of //8 or higher removes the 

April, 1941] 



Fig. 4 shows the attachment with the slate-box removed and with 
the operating lever carrying the auxiliary optical system blocked in 
its "raised" position. The hinged cover on the front surface of the 
attachment housing carries two of the 2V2-volt, 1 / 2 -ampere bulbs 
which illuminate the slate-box face. The housing attached at the 
top of the vertical arm carries the sub-assembly consisting of the first- 
surface mirror and the two focus-correcting lenses. 

Fig. 5 shows the attachment with 
the operating lever in its raised 
position and with one of the lamp 
covers swung open to show the 
ease of bulb replacement. The 
bulbs in their sockets are shown 
removed from their reflector, which 
is integral with the lamp cover 
casting. The separate piece shown 
to the left is the clamping plate 
which holds the bulbs and their 
bayonet sockets in position. It is 
held in place by a thumbscrew. 
The face of the slate-box may be 
seen at the end of the housing. 

Fig. 6 shows the slate-box face. The white characters on the black 
background are on manually operated, numbered wheels which are 
adjusted to show the scene and take numbers. The first wheel is 
provided with letters to identify the following special types of takes : 

E t Special Effect 

R, Retake 

r, Transparency (Process) 

Z, ' Test 

FIG. 6. 

View of slate-box face 
information carried 

This first wheel contains also the digits 1, 2, 3, etc., which permit 
setting up scene numbers as high as 3999 should such a necessity 
arise. The next three wheels each carry the numbers to 9, plus 
blanks, and constitute the scene numbers. 

The fifth wheel bears the following notations : 

X, Sync (wild motor) Shot 
A, B, C, D, Camera Angles 
Ax, Bx, Cx, Dx, Sync Shots from Indicated Angles 



LT. S. M. p. E. 

The remaining two wheels carry numbers from to 9 plus blanks, 
and are used to number takes. Thus the figure as shown in Fig. 6 
represents Scene No. 999, which is a transparency set-up, take No. 
99, made without sound (sync), and from 
camera angle D. 

In order to provide a degree of flexi- 
bility as to the information carried by 
the slate, three removable, white, matte 
Vinylite strips are provided upon which 
notations may be made in pencil or printed. 
The information normally required in addi- 
tion to that provided by the numbered wheels 
is as described on the white inserts of Fig. 6. 

The slate-box with all data may be removed readily from the slating 
attachment housing for the insertion of new slips, or for alteration of 
the date. It is not necessary to remove the slate-box to set scene 

FIG. 7. Battery-box. 

FIG. 8. Attachment and battery-box shown mounted on NC 
blimp (sunshade removed) . 

and take numbers. The seven numbered wheels extend through the 
rear of the slate-box and carry a second set of numbers visible to the 
operator at all times and indicating the corresponding data appearing 
on the slate face. 


Fig. 7 shows the battery -box. It weighs 5 l /2 pounds and is 6 
inches long, 3 inches wide, and 6 3 /4 inches high. The three-position 
selector switch located beneath the strap handle adjusts the slate 
illumination. The switch plate is marked in three steps "//2.3-4," 
"//4-8," and "//8-16." The switch is set to include the lens stop in 
use for a particular take. On the setting including //1 6 satisfactory- 
exposure is obtained for a stop of //22 although a condition of under- 
exposure is being approached for certain picture printing conditions. 

FIG. 9. Attachment shown mounted on open tripod set-up of 
NC camera. 

The batteries used are of the 6- volt radio A -battery type connected 
in parallel. The battery capacity has proved adequate for from two 
to three months' service in normal use on studio sets. 

Fig. 8 shows the attachment and battery-box mounted upon the 
NC camera blimp, in the operating position. The sunshade has been 
removed from the blimp to afford a better view of the slater 

As the attachment arm reaches its stop, the arm actuates a micro- 
switch connecting the power supply to the bulbs which illuminate 
the slate face. When the operator removes his thumb from the 
operating lever, the arm swings down of its own weight into the at- 

364 F. C. GILBERT [j. s. M. P. E. 

tachment housing and the switch opens, turning off the lights. The 
arm is made so as to furnish a reasonably dustproof cover for the 

The batteries are connected to the attachment by a cable and plug. 
The plug is equipped with a latch which must be released by hand be- 
fore the plug can be separated from the attachment. A second 
latch, released by the insertion of the connecting plug, locks the slater 
arm in its closed position, thereby preventing slating unless the 
battery connection has been made. 

FIG. 10. Slater-equipped NC camera, on 

The sunshade has a slot in its bottom surface through which the 
slater arm is raised into its operating position. 

Fig. 9 shows the attachment in its closed position mounted upon 
an open tripod or "sync" set-up of the Mitchell NC camera. It af- 
fords a view of the knurled wheels and tell-tale numbers on the rear 
face of the slate-box. When the 100-mm camera lens is employed in 
this sync set-up, it is necessary to operate with the sunshade removed. 

Fig. 10 shows the slater in use on an open Mitchell NC camera 
in the air on one of our large booms in which situation its ad- 


vantages are obvious. Where the old slating method was employed 
such situations resulted in slate images so small as to require the use of 
a magnifier or projection of the picture to identify the take. Addi- 
tional camera situations, many encountered in daily practice, where 
the slating attachment is a distinct advantage, will undoubtedly sug- 
gest themselves to those familiar with the various phases of motion 
picture photography. 

The first slater of this type was placed on field trial in production 
use early in January, 1940. All cameras in use on the lot and 
on location are now equipped with the attachment as described. 
After several months of such general use we find the device and the 
slates it produces popular with the various technicians whose work 
required their use. 



Summary. This work was undertaken to determine the benefits of a true mono- 
chromatic optical system for variable-density recording in the ultraviolet region. A 
full quartz optical system consisting of both spherical and cylindrical lenses was used, 
hating a reduction of 10-1 from the light-valve spacing. The reduction in lens dis- 
tortion and improvement in general image quality is reported along with intermodula- 
tion tests on the system, which uses an automatic air -controlled mercury-vapor lamp 

Much emphasis has been placed on the necessity of reducing 
modulation products in a film-recording system. Previously re- 
ported analyses indicated the advantage to be obtained by reduction 
of the effective height of the image in variable-density recording with 
a light-valve. 

Fig. 1 shows the computed intermodulation due to ribbon velocity 
effect when 60 cycles and a higher frequency are recorded simul- 
taneously on the film with a 12-db differential between the two fre- 
quencies. 1 From these data it will be noted that the height of the 
mage on the film should never be greater than 0.25 mil, and it is pref- 
ierable to reduce the image still more. 

Originally the light-valve ribbons were spaced 2 mils apart, using 
a 2 to 1 optical reduction on the film ; very soon after this, the spacing 
was reduced to 1 mil. This reduced the effect of high-frequency 
loss and intermodulation considerably, but the need of further im- 
provement was recognized. 

Several years ago, in an effort to make a variable-intensity system, 
a double quartz wedge was inserted between the light-valve and the 
film so that the variable width of the light-valve was changed into 
variable-intensity at the film. This system reduced the intermodula- 
tion at high frequencies to a very low degree, but its commercial ap- 
plication was retarded due to its poor light efficiency. Later, a 4 to 

* Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received Novem- 
ber 19, 1940. 

** Metro-Goldwyn-Mayer Pictures, Culver City, Calif. 

& The Society is not responsible for statements by authors & 



1 optical system was made available such that a 0.25-mil image of a 
1-mil light- valve was set up on the film with approximately 50 per 
cent loss of light. This system is now in general use today in prac- 
tically all release recording machines and in a great number of original 
negative production machines. 

However, due to the fact that the current standard practice uses 
1.5 to 2-mil movements of the ribbons, the image on the film is as 





A -0.5 


80 <*> 
80 <to 



I 20 

FIG. 1. 

Computed intermodulation due to rib- 
bon velocity effect. 

great as 0.5 mil. This image width at high levels produces com- 
paratively high intermodulation products and the need of reducing 
these products still more is very desirable. 

Improvements due to further correction of spherical and chromatic 
aberration in recording optical systems has indicated the need for 
the reduction of this type of distortion to a minimum. It has been 
felt that a true monochromatic lens system in which a nearer ap- 
proximation to variable-intensity could be accomplished would re- 
move to a minimum all the distortions pointed out above. 


O. L. DUPY AND J. K. MILLIARD [j. s. M. P. E. 

The improvement and practical application of the mercury arc 
as a source of illumination for variable-density recording encouraged 
the development of the monochromatic system, since a light-intensity 
many times that which could be obtained from the incandescent 
lamp is available in the spectral band desired. 

Previously published work has shown the bad flare on ordinary posi- 
tive stock when exposed with incandescent light at both high and 
medium low frequencies due to scattering within the emulsion and 
due to reflection from the back surface of the base into the emulsion. 
The use of filtered light, or the mercury arc, as a source materially 






FIG. 2. Monochromatic optical system for variable-density re- 

reduces these effects. This is due to the fact that the absorption of 
ultraviolet radiation by the emulsion is very great. 2 

For this reason it was decided to set up a recording optical system 
where the exposure was confined to the single 3650 A spectral line. 
This system is composed of a mercury arc light-source, ultraviolet 
filter, and a full quartz spherical condenser objective lens system, 
each element having a focal length of 3 inches. The optical system 
was further modified by providing a very small quartz cylindrical 
lens spaced 20 mils from the film and having a reduction of approxi- 
mately 5 to 1 in the vertical direction. 3 ' * 

* Hardy has published information 3 on the use of the cylindrical lens in 
film-recording systems to change the magnification in one direction without 
change in the other direction. 



Since this cylinder permits focusing the slit in the direction parallel 
to the slit, the rays that are at right angles to the plane of the slit 
must necessarily be focused at the plane of the film. This is ac- 
complished by a cylindrical lens with axis perpendicular to the slit 
and placed immediately in front of the objective. This, then, gives 
an overall reduction of approximately 10 to 1 in the plane of the slit 
and a 2 to 1 reduction in a plane at right angles to the slit. Fig. 2 
shows the schematic arrangement of the complete optical system. 

Since the illumination is provided principally from the 3650 A mer- 
cury line, a simple quartz optical system may be used. This reduces 

FIG. 3. Photomicrograph of sound-track made with the optical system 
described, when modulated at 14,000 cycles (reduced to 0.73 the size of the 
original photograph) . 

the need for chromatic correction in the lens system. It is recognized 
that quartz has the highest transmission in the range. Therefore 
the overall light efficiency is comparatively high and its supply is 
commercially available. 

With this set-up the overall track illumination is much more uni- 
form than has heretofore been secured by means of tungsten filament 
lamps and spherical lens only. As a result, when a push-pull track 
is made, the cancellation products are much improved at the higher 
frequencies, especially in the case of the wide (200-mil) push-pull 
track. Normally in corrected lens systems available today, the 
maximum cancellation obtained from a push-pull track when placed 
pn a regular reproducer is 25 to 30 db at 1000 cycles and somewhat 

370 O. L. DUPY AND J. K. MILLIARD [j. s. M. p. E. 

lower at 7000 cycles. With the above-described system the can- 
cellation is not less than 30 db for all frequencies over the entire range. 

With this system, using the controlled air motor on the mercury 
arc lamp, 4 the density variation on the track is considerably more 
uniform than that which can be obtained with tungsten light. In 
push-pull recording the track balance between both halves is very 
important and laboratory daily density reports indicate balances 
within 0.02 over long periods of time. Fig. 3 shows a photomicro- 
graph of a sound-track made with this optical system when modulated 
at 14,000 cycles. 

The intermodulation terms obtained from this system due to 
ribbon-velocity effects are lower than those reported from any other 
system now being used in commercial production. Measurements 
using fine-grain negative and prints indicate less than 4 per cent 
intermodulation (approximately 1 per cent total harmonic content) 
as an optimum over a wide range of printer lights. Listening tests 
verify the degree of improvement by reducing to a minimum the 
effects of chromatic aberration, reduction of intermodulation, and 
non-uniformity of track illumination. It is believed that for the 
first time a variable-density recording system has been used in which 
a true monochromatic optical system in the ultraviolet range has 
been applied in commercial production. Valuable work was done by 
Mr. J. W. Stafford in the design and testing of this optical system. 


1 FRAYNE, J. G., AND SCOVILLE, R. R.: "Analysis and Measurements of Dis- 
tortion in Variable-Density Recording," /. Soc. Mot. Pict. Eng., XXXII (June, 
1939), p. 650. 

2 FRAYNE, J. G., AND PAGHARULA, V. : "The Effects of Ultraviolet Light on 
Variable-Density Recording and Printing," J. Soc. Mot. Pict. Eng., XXXIV (June, 
1940), p. 614. 

3 HARDY, A. C.: "The Optics of Sound-Recording Systems," Trans. Soc. 
Mot. Pict. Eng., XII (Sept., 1928), No. 35, p. 760; see also J. Opt. Soc. Amer., 14 
(1927), p. 505. 

4 DUPY, O. L., AND BILLIARD, J. K.: "Increased Illumination for Fine-Grain 
Recording," Amer. Cinemat. (Jan., 1940), p. 36. 


DR. FRAYNE: I notice that Mr. Dupy is using a l-mm 584 Corning filter. 
That filter does not make this a 100-per cent monochromatic system because it 
lets through about 25 per cent of the line at 4000 A. I wonder whether he has 
made any test with the 2-mm 584. 


MR. DUPY: An increase in the amount of filtering gives us only a reduction 
in light, and the light is sufficiently monochromatic to permit the use of an optical 
system lacking chromatic correction. 

DR. SANDVIK: You might possibly get better image definition by using a 
molten lens. In that way it is possible to get lenses that are not cylindrical or of 
circular section. They are made by grinding a glass in a huge molten, and take a 
very fine polish. You could not use quartz, however, because quartz is too hard 
for the purpose. If you used a ground glass you would not get enough absorp- 
tion at 3650 A. 

MR. DUPY: That sounds interesting. At present we have a source of light 
and a cheap optical system that seem to stay with us. They are not perfect by 
any means, but form a tool that will at present work up to the imits of our knowl- 
edge of how to use them. 

MR. HILLIARD : There seems to be an impression that the incandescent lamp 
is the most reliable source available to date. In our experience with the mercury- 
vapor lamp over a year's time, some of the lamps have stayed in the recorder as 
long as six months. With the motor control blower it is possible to produce 
variable-density film, or any exposure for that matter, with less deviation than 
has normally been recognized in the incandescent lamp system. 

We have had as much as an 0.05 change in density on a negative with approxi 
mately 0.06 density; whereas over the period of a year's time we have been able 
to cut that down to approximately 0.02 with the controlled mercury-vapor light. 
For those who are interested in the laboratory part of the work it would appear 
profitable to utilize this function so as to obtain a controlled density in the re- 
leased printing by this method. The laboratory people have felt that the in- 
candescent lamp was the most foolproof, infallible, as well as the most uniform 
source, and I certainly would like them to consider the mercury lamp as such a re- 
placement for the uniformity which they now have. 

MR. LAUBE: Have you found it practicable to install mercury lamps in in- 
stalled channels. 

MR. HILLIARD: It is more practicable to install them in fixed channels, be- 
cause there is some difficulty in using such a system with the regulated air-steam, 
as compared to the incandescent lamp. But in the case of release machines and 
in the laboratory, which is a fixed condition, there should be no objection to it, 
outside the fact that it does look slightly more complicated on the surface, and 
you may perhaps have to spend a little more money in the beginning to put it into 
operation. But it does give the possibility of obtaining a result in excess of what 
we are able to get at the present tune, namely, the exposure of a fine-grain re- 

MR. LAUBE : I understand that a feedback circuit has been devised that will 
keep the intensity of the lamp within very small fluctuations, 

MR. HILLIARD: The motor-controlled blower is nothing but a mechanical 
feedback similar to the electrical feedback we use in amplifiers. 

Equilibrium occurs when there is no change in exposure up to the degree of 
feedback that is allowed. When you get outside the limit, which is a matter, say, 
of 2 :1 in exposure, then there is some exposure change. But where it is a matter 
of 5, 15, and 20 watts' change, there is more than enough feedback in the system 
to take care of the irregularity. 

372 O. L. DUPY AND J. K. MILLIARD [J. S. M. P. E. 

MR. LAUBE: I meant that the feedback principle may be used also with in- 
candescent lamps. 

MR. HILLIARD: I understand. I simply wanted to point out that we know 
what the limitation of melting tungsten is, as regards exposure, and the mercury- 
arc lamp does offer a source of intensity beyond that point. 

MR. ALBIN: In the monochromatic system using quartz throughout, what 
was the effective change of gamma as compared to the ordinary glass incandes- 
cent source? 

MR. MILLIARD : Are you speaking of the negative, or the overall effect? 

MR. ALBIN : Everything but the negative. 

MR. MILLIARD: Starting with a release negative gamma of approximately 0.4, 
it has been fairly well established that a reduction of 10 per cent can be obtained 
if the exposure is limited to the region around 3650 A. 

MR. ALBIN : How about the negative ? 

DR. FRAYNE : On the negative we found, for a 584 filter, a 15-per cent reduc- 
tion in the negative gamma as contrasted to about 40 per cent in the positive. 

MR. MILLIARD: It depends upon the filtering. I have gone down to a 3-mil 
584 filter, and we estimate roughly a 15-per cent reduction at that point, which I 
think was very drastic. 

MR. ALBIN : If this negative were used for release, it would be necessary to in- 
crease the negative gamma, that is, the developed gamma of the negative, by that 
percentage, and would tend to exaggerate any variation of density due to variation 
of exposure. 

MR. HILLARD : That is true. But you do get something else for your tendency 
to exaggerate, do you not? 

MR. ALBIN: Yes. 

MR. MILLIARD: It is a balance of one against the other, and depends upon 
which is the worse. I think we all recognize that the monochromatic element, 
both in printing and in exposure, has several decided advantages. 

MR. ALBIN: But it merely speaks that much more for your control, the 
fact that you get less variation even with the higher gamma. 

MR. MILLIARD: It depends upon our method of interpreting the gamma. A 
unity gamma is whatever the negative might be printed up in a reciprocal manner 
so as to get an overall unity gamma, is it not? 

MR. KELLOGG: Mr. Milliard, about how much change in light, as determined 
photographically, does a given percentage change in current produce, assuming 
that the change takes place so quickly that the temperature is constant. For 
example, if there is a sudden increase in current amount to 2 per cent, by what 
percentage would the light increase? 

MR. MILLIARD : In the case of a sudden change of 2 per cent in current, assum- 
ing it changes so quickly the temperature is constant (which means the voltage is 
constant), the wattage increase is E X (/ X 1.02). Since the illumination is 
directly proportional to wattage, the illumination will change by 2 per cent. This 
corresponds to a density change of approximately 0.01 at a gamma of 0.42. 

MR. KELLOGG : Did I understand Mr. Milliard to say that he does not have a 
regulated supply? 

MR. MILLIARD : Yes. When the arc voltage increases, the slower motor, which 
is tied across the arc terminals, also speeds up and forces more air through, which 


decreases the arc voltage and tends to hold the illumination constant in spite of 
varying line voltage. 

MR. KELLOGG : Has that been satisfactory with various power sources? 

MR. HILLIARD : It has seemed to be. In the daily logs sent up by the labora- 
tory there has been no reported deviation in excess of 0.02 to 0.03 in density at 
the nominal 0.6 normal density. 

MR. KELLOGG : Do you have to filter the supply pretty well ? 

MR. HILLIARD: Yes, to get the ripple out, because the instantaneous effect of 
the mercury light would record naturally any 50 or 100-cycle components that 
might be in it. There is considerable inductance and capacity to give a "fly- 
wheel" effect. 



Summary. The observations reported are the result of an investigation to de- 
termine the effect of a delay between the exposure and development of modern motion 
picture films. The stability of the latent image in terms of speed, gradation, graini- 
ness, and color response has been studied. 

In general, a definite speed increase was noted on negative emulsions, a decrease 
on positive emulsions. There were also changes in gradation and graininess. The 
detailed findings which vary considerably with the individual emulsion type are 
given, followed by a general discussion and interpretation of the results. A brief 
review of the literature is included. 

A little over a year ago we started an investigation of the effect of 
age on the photographic latent image. This, prompted by a desire 
to determine the effect of a delay between the exposure and develop- 
ment of motion picture films, embraced cine positive, infrared, and 
three speed ranges of panchromatic negative films. 

There are, in the photographic trade, many fixed ideas on the result 
of delaying development after exposure, most of them contradicting 
each other. In view of these conflicting beliefs it is surprising that so 
few actual investigations have been made and reported in the litera- 
ture. It is even more surprising that such a small portion of those 
that did appear were well organized and yielded valid results. 

Our literature search uncovered many articles based either upon 
unsubstantiated opinions or incomplete and misleading investiga- 
tions. In fact, this led E. Heisenberg 1 to remark, "It is striking 
that in almost all the older investigations only a regression of the 
latent image has come to the foreground." We shall mention only a 
few of these older investigations in passing and devote our time to 
the more careful research work done by Bullock, Heisenberg, Jaus- 
seran, Mees, and others. 

As long ago as 1889 Bedding 2 called the continuing action of light 

*Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received October 5, 

**Agfa Anco Corp., Binghamton, N. Y. 


$ The Society is not responsible for statements by authors & 


after exposure, "a pretty little photographic heresy." On the other 
hand, W. I. Rogers 3 mentioned in 1891 that plates known to be under- 
exposed can be intensified by delaying development for several weeks. 
C. E. K. Mees 4 in 1915 observed a speed increase if development 
of the latent image was delayed, and G. Jausseran 5 discovered in 
1929 a linear relation between the logarithm of the delay time be- 
tween exposure and development and the resulting density. This 
was confirmed by E. R. Bullock 6 in his very thorough and extensive 
work on the subject. 

Various reports on factors causing or influencing changes in the 
latent image have appeared. J. Barker 7 found bromide emulsions 
more stable than iodide or iodide-bromide emulsions. Bakeland 8 
reported greater stability of neutral or weakly alkaline emulsions 
compared to acid emulsions. According to Heisenberg 1 the growth 
of the latent image on panchromatic emulsions is due to the sensitiza- 
tion, and according to Bullock 6 it varies with the development 
gamma and the speed of the emulsion. The wavelength of the ex- 
posure light and the strength of the developer are also factors. 

These investigators were guided principally by theoretical interests 
and were not concerned with the practical significance of latent image 
intensification. Their investigations were for the most part concerned 
with deriving a law defining the latent image intensification or re- 
gression with time, which would allow this phenomenon to be inter- 
preted in terms of the latest theory explaining the physico-chemical 
nature of latent image formation. 

On the other hand, the tests which are to be reported here were 
carried out under average working conditions to obtain data that 
would interest the motion picture camerman and film technician. 
They are, it is true, more qualitative than quantitative, but permit 
valid conclusions to be drawn. 

In our investigation the effect that aging the latent image has on 
the speed, gradation, general shape of the ZMog E curves, graininess, 
and color response of the developable image was studied. Our re- 
sults may be used to answer such practical questions as : How well do 
various films keep after exposure but before development? What 
changes in speed and gradation take place when development is 
delayed and are they sufficiently great to merit serious consideration? 
Is special exposure and development necessary to obtain the best 
results when it is known at the time of exposure that there will be 
a considerable delay before development? 

376 K. FAMULENER AND E. LOESSEL [j. s. M. p. E. 

We selected for this work widely used motion picture films manu- 
factured in this country. They were representative of the high-speed, 
medium-speed, and low-speed panchromatic and infrared negative 
films as well as motion picture positive. The procedure which we 
worked out was to cut from each film under investigation 104 4-ft 
35-mm strips. Half of these were exposed on a time-scale sensitome- 
ter at the beginning of the study. The other half were exposed 
on the same apparatus once each week during the investigation. 
Two butted exposures with the lamp and filter appropriate to the 
particular film were made on each strip. Once each week one of the 
originally exposed strips and, a freshly exposed strip were developed 
simultaneously on a developing machine using the developer and 
time of development most suitable to the emulsion being studied. 
Since the strips were always developed together variations in develop- 
ment were cancelled out. Both the exposed and unexposed strips 
were stored during the course of this investigation in taped cans at 
a temperature of approximately 70F and relative humidity of 50 
per cent. The films were handled both before packing and during 
exposure under these same conditions of temperature and humidity. 
After development the strips to be compared were read, averaged, 
and their ZMog E curves plotted. These plots furnished us with the 
basic sensitometric data. 

Practical camera tests were made in addition to the sensitometric 
measurements. A set of pictorials, exposed in a Leica camera at 
Vz-stop intervals, was made on the first part of each of two 36- 
exposure cartridges at the beginning of the investigation. A dummy 
test set was used and the position and intensity of lights, camera 
position, etc., carefully measured and recorded. A photoelectric 
photometer gave a rough value for the illumination level. This 
enabled us to duplicate the set for later tests. Four months after 
the beginning of the investigation another series of exposures were 
made on one of the cartridges and eight months later on the other. 
These were machine developed immediately following exposure of 
the second cartridge. Thus pictorial comparisons were prepared of 
film developed immediately after exposure and the same film de- 
veloped four and eight months, respectively, thereafter. Due to 
the looser control of the duplication of the pictorial tests, the results 
obtained from these have more a corroborative than absolute value. 
An Agfa step color-chart was used in these pictorials to give us an 
indication of any change in the color response values. 

April, 1941] 



The results of the investigation are shown in the figures illustrating 
the detailed discussion. We have, plotted on one chart, curves for 
development, 0, 13, 26, 39, and 48 weeks after exposure. These were 
obtained by averaging the curves representing the film developed 
immediately after exposure and relating the other curves to this 
average. The detailed results which this data yielded were as fol- 

High-Speed Panchromatic Negative. In Figs. 1 and 2 are shown the 
curves obtained on two competitive high-speed motion picture 
negative films. In Fig.l it is seen that as the development of the 


13 . 

2.6 . 

REL Loo E 

.6 .9 


Z.I Z.I- *-7 

FIG. 1. Latent image age effect on high-speed panchro- 
matic motion picture negative film no. 1. 

film is delayed longer and longer after exposure, the apparent speed 
of the film increases. This increase is, during the first four weeks, 
slight and general for the entire characteristic curve. After five 
weeks of delay the film begins to show a preferential speed increase 
at densities around 0.6. This continues with age and constantly ex- 
tends further into the toe section. When 39 weeks have elapsed the 
characteristic curve has lost all resemblance to the curve obtained 
from the same film freshly exposed and developed. There is a speed 
increase of 2*/2 stops at a density of 0.6 above fog and above this 
value the film is flatter, and below, steeper than the same material 
developed immediately after exposure. Thirteen weeks later the 
curves show little additional change. Pictorial tests developed four 



[J. S. M. P. E. 

and eight months after exposure demonstrate a speed increase in the 
toe region of l x /2 and 2 stops, respectively. In both cases the pic- 
torial gradation was lower than that of the same film exposed and 
developed without delay. 

Grain enlargements made from both pictorial and sensitometric 
strips indicate an increase in graininess with delayed development. 
The color-sensitivity from step color-chart readings shows a drop in 
blue and green relative to the red and yellow. 

The changes in the sensitometric characteristics of the film shown 
in Fig. 2 differ from those found on the film represented by Fig. 1. 






13 ' 


REL. Lo& E 

.3 .6 



FIG. 2. Latent image age effect on high-speed panchro- 
matic motion picture negative film no. 2. 

There is a rapid increase in speed, a full stop being noted after only 
three weeks. Six weeks after the exposure, the speed increases I 1 / \ 
stops and there is no appreciable change in speed for longer intervals 
between exposures and development. There is a slight decrease in 
contrast with delayed development. Enlargements made from the 
sensitometric strips showed a barely preceptibly greater graininess 
on the film which was exposed 48 weeks before development compared 
with that exposed immediately before development. 

Thus for high-speed panchromatic films there is an increase in 
speed of from l x /2 to 2 l / 2 stops, a decrease in gradation and a slight 
increase in grain. 

April, 1941] 



Medium-Speed Panchromatic Motion Picture Negative Films. The 
films that we have classed as medium-speed panchromatic negative 
motion picture emulsions are those which each of the leading Ameri- 
can film manufacturers offered two years ago as a standard motion 
picture negative material. It is these which were used for all routine 
studio production work. 

The greatest speed increase was noted on the film shown in Fig. 
3 which was found to be P/2 stops faster with a 13- week delay and 2 
stops faster with a 39-week delay than comparison strips developed 
immediately after exposure. The characteristics appear to be stable 




13 . 


REL. Loo E 

O .3 .(, .9 1.2 I.S /.O 2.1 24 Z7 3O 

FIG. 3. Latent image age effect on medium-speed pan- 
chromatic motion picture negative film no. 3. 

after 39 weeks, no significant change being noted after 13 more weeks 
of aging. 

The film shown in Fig. 4 exhibits a faster latent image growth than 
that in Fig. 3 and little change after 13 weeks save for growth in the 
shoulder densities. There appears to be no change whatsoever after 
36 weeks' aging at which time the speed increase is slightly more than 
l l /z stops. 

In Fig. 5 we noted a similar effect, the film gaining 1 stop in speed 
during the first 13 weeks, l x /2 stops during the first 26 weeks and 
showing little change thereafter. Pictorials bear out the sensito- 
metric results obtained on these three films. In all cases there is a 
slight coarsening of grain with and little change in color re- 



[J. S. M. P. E. 

sponse. Aging of the latent image brings about a pronounced soften- 
ing in highlight gradation on the film shown in Fig. 3 and a very 
slight softening on those shown in Figs. 4 and 5. 

Fine-Grain Panchromatic Motion Picture Negative Film. Under 
this classification only one film was tested. The results are shown in 
Fig. 6. Here latent image age resulted in a speed increase of 1 / 2 
stop in 13 weeks, l x /2 stops in 39 weeks, and 2 stops in 48 weeks. 
There is a definite decrease in gradation in the high densities above 
1.0, a steepening in the lower densities. Pictorial tests showed an 
increase in speed of 1 to iy 2 stops with a loss in highlight gradation. 



f !< 



3.0 3.3 

FIG. 4. Latent image age effect on medium-speed panchromatic 
motion picture negative film no. 4. 

There was a definite coarsening of grain with the age of the latent 
image which will be discussed later. Color-chart readings showed a 
drop in blue-green response compared with the red-yellow. 

Infrared Negative Motion Picture Negative Film. One infrared 
motion picture film was tested with the results shown in Fig. 7. It 
will be noticed that there is a slight increase in speed and a very slight 
loss in gradation with latent image aging. The speed increase is of 
the order of y 2 stop and there is little change after 13 weeks save 
an increase in toe speed. 

Cine Positive Emulsions. Two motion picture cine positive emul- 
sions were tested and the results plotted in Figs. 8 and 9, respectively. 

April, 1941] 



In Fig. 8 we find that there is very little change in speed or grada- 
tion with latent image aging. Matched densities from strips exposed 
immediately before development and strips exposed one year before 


l ,. . 





FIG. 5. Latent image age effect on medium-speed panchromatic 
motion picture negative film no. 5. 

13 " 


T?E:L. LOG- E 

FIG. 6. Latent image age effect on fine-grain panchro- 
matic motion picture negative film. ', . 

development were projected side by side and no appreciable grain 
noted on the screen. The effective enlargement was about equal tQ 
that obtained in the average motion picture theater, 



[J. S. M. p. E. 

In Fig. 9 there was a marked decrease in speed with age amounting 
to approximately 7 printer points in 13 weeks and rising to about 15 
printer points in 39 weeks. Grain projections showed no discernible 
difference between the aged and freshly exposed film. The results 
on the two films are best shown in Table I. 


Image Age 

Film No. 
Change in 




Vy si* - 



Vysl + 



Vysl + 



2 points -f- 


Film No. 9 
Change in 

7 + 
12 + 
15 + 
15 + 



*"Vy si" = very slight. 


. Loo- E 

-3 .6 .9 1.2 US X8 XI 2+ *7 3.O J.j 

FIG. 7. Latent image age effect on infrared motion picture nega- 
tive film. 


Intensification or Regression of the Latent Image with Age. The 
test results obtained under the different emulsion groups indicate 
that the intensification or regression of the latent image varies with 
several factors. First, we find, as Heisenberg contended, that the 
intensification is most pronounced with color sensitized emulsions. 
All the sensitized negative materials show a speed increase of from 

April, 1941] 



3 /4 to 2 J /2 stops. The two positive films which, of course, were not 
sensitized, showed no change and a decided regression of the latent 
image, respectively. Furthermore, it is apparent that the greatest 
intensification of the latent image is to be found on films with a high 
initial sensitivity. The high-speed films showed an increase in speed 
of approximately 2 stops, medium-speed I 1 /2to2, and the slow-speed 
films 1 stop, while with infrared there was a 3 /4 of a stop increase. 
This confirms Bullock's claim that the degree of latent image intensi- 
fication varies with the initial emulsion speed. 



3 .6 .9 /.z /.s /e z.i z.+ 2.7 3.0 
FIG. 8. Latent image age effect on motion picture positive 
film no. 8. 

Bullock noted that the advance in speed is most pronounced at 
densities near 0.6 which was confirmed on several films. However, 
our tests showed sufficient exceptions for us to conclude that this 
initial increase in speed in the neighborhood of 0.6 is not a necessary 
accompaniment of latent image age. 

The shoulder of the characteristic curve does not show as marked 
an image growth as do the lower densities. In some cases, in fact, a 
regression may occur at extremely high densities. This indicates 
that the intensification or growth effect depends upon the exposure 
and may reverse itself at high levels. The densities at which no 
growth or decay takes place, i, e., at which the two characteristic 



[J. S. M. P. E. 

curves intersect, is an ideal point at which to make comparisons of 
other age effects. 

Jausseran derived a logarithmic relation between the intensification 
and time lapse between exposure and development. This was later 
confirmed by Bullock. However, the ages considered were in units 
of seconds rather than weeks, the longest interval between exposure 
and development being five hours. Bullock implies further that the 
growth effect does not continue long and says, "Since the rate of 

.3 .6 .9 /2 AS /O 9-1 2.+ 2-7 

FIG. 9. Latent image age effect on motion picture positive 
film no. 9. 

increase of the latent image may. . . .be taken as inversely pro- 
portional to the time, the practice of keeping exposed material for a 
few hours or overnight whenever it is desired to develop non-simul- 
taneous low intensity exposures simultaneously, is to be commended." 
Our work extending over periods of months rather than hours showed 
a latent image growth far beyond the "overnight" period that 
Bullock mentioned. We can not establish a definite law for this 
growth but in general it approaches a logarithmic function. 

The effect is not, as Bullock states, general for all emulsions, since 
the positive emulsion discussed under Fig. 9 shows a definite regres- 
sion, However, again we must remember that Bullock speaks in 

April, 1941] 



terms of hours and we in terms of months, and it is possible that there 
is an increase of the developable image for a short time after exposure 
followed by a regression. This regression is not even necessarily 
related to the growth or intensification effect. In fact, at no time in 
our work did we observe the growth effect followed by regression of 
the developable image. On the contrary, the growth effect reached a 
maximum which was maintained throughout the balance of the test. 
The time required to reach this maximum varied with the individual 
emulsion from a few weeks to 9 months but was not characteristic 


(/) SO- 


? 60- 



cr 4 OH 




FIG. 10. 

0.2 0.4 0.6 0.8 LOO 1.20 1.40 

Graininess-density curve for high-speed panchromatic 
motion picture negative film no. 1. 

of any particular emulsion grouped according to speed or sensitiza- 
tion. In comparing Bullock's results with ours, it should be remem- 
bered that besides the difference in magnitude of the intervals 
studied he used an intensity-scale exposure method while we used a 
time-scale method. 

Effect of Image Growth on Grain. Our investigations showed that 
in many cases the growth of the developable image with age was ac- 
companied by an increase in graininess. In order to get a better 
approximation of the difference in graininess than that afforded by 
grain enlargements, use was made of the graininess meter described 
by Goetz. 9 The graininess-density curve for the higher speed pan- 
chromatic negative film, whose sensitometric curves are given in 


Fig. 1, is shown in Fig. 10. From Fig. 1 we note that there has been 
no image growth at a density of 1.20. However, the density graininess 
curves do not approach each other at this density. Therefore, it 
appears that while the developable image becomes more grainy with 
age, this increase in graininess does not account for the increase in 

Color-Sensitivity and the Growth Effect. LeClerc in "Photography: 
Theory and Practice" says that Jausseran found that the growth ef- 
fect is not the same for latent images produced by different radiations. 
In our work no more accurate check was made on the dependence of 
the growth effect on the spectral quality of the original exposure than 
was afforded by the Agfa step color-chart in pictorial tests. This 
indicates that there was less growth in the blue and green than in the 
yellow and red range on panchromatic emulsions. However, further 
work is being done using more accurate methods. One would expect 
less growth effect with a blue light exposure since it has already been 
established that growth of the developable image is closely related 
to the optical sensitization of the emulsion. Our results with un- 
sensitized emulsions, such as motion picture positive film, show no 
growth effect in one case and a decided regression in the other. 

Mechanism of the Growth Effect. We do not feel that our data are 
sufficiently refined at the present time to permit conclusions on the 
mechanism of the growth effect. However, we have found in this 
investigation some facts that substantiate and some that contradict 
the theories that have been advanced to account for the phenomenon. 
We do not wish to discuss these until further more accurate work has 
been done. 


This investigation may be summarized as follows : 

(1) This investigation differs from those of previous workers by 
tracing the change of the developable image over a period of a year 
rather than for several hours after exposure. 

(2) A growth of the latent image varying from x /2 to 2 x /2 stops 
was found in all the negative films tested. 

(5) No intensification in one case and a decided regression in 
another was found with cine positive films. 

(4) The growth effect depends upon the sensitivity of the emul- 
sion, is more pronounced in fast than in slow emulsions. 


(5) The intensification effect is most pronounced at low densities, 
and may be entirely lacking or replaced by a regression at high densi- 

(6) The growth effect depends on the optical sensitization of the 
emulsion. Unsensitized emulsions show little growth effect, some- 
times decided regression. 

(7) The growth effect results in increased graininess but this 
increase is not sufficient to account for the gain in density. 

() The spectral quality of the exposure illumination may in- 
fluence the growth effect. This is suggested by the present work but 
requires a more accurate method for verification. 


1 HEISENBERG, E. : "Die Veranderung des latenten Bildes in Halogensilber- 
gelatineschichten bei der Lagerung," Veroff. des wiss. Zentral-Laboratoriums der 
photographischen Abteilung, AGFA (I. G. Farbenindustrie Aktiengesellschaft), 
3, S. Hirzel, Leipzig (1933), p. 47. 

2 BEDDING, T.: "Continuating Action of Light," Brit. J. Phot., 36 (Sept. 20, 
1889), p. 619. 

3 ROGERS, W. I.: "Casket of Photographic Gems," London (1891), p. 35. 

4 MEES, C. E. K.: "A constate un accroissement de sensibilite," Phot, and 
Focus, 39 (May 11, 1915), p. 338. 

6 JAUSSERAN, G.: "Constation et etude quantitative du renforcement spon- 
tane," Rev. d'Optigue, 8 (March, 1929), p. 119. 

6 BULLOCK, E. R. : "Sur le renforcement spontane de 1'image latente entre la 
pose et le developpement," Science et Industries Photographigues (May, 1930), 
p. 169; (Jan., 1933), p. 6; (Feb., 1933), p. 32. 

7 EDER, J. M. (Editor) : "Jahrbuch fur Photographic und Reproduktions- 
technik fur das Jahr 1905," Wilhelm Knapp, Halle a. S., 19 (1905), p. 364. 

8 EDER, J. M. (Editor) : "Jahrbuch fur Photographic und Reproduktions- 
technik fur das Jahr 1904," Wilhelm Knapp, Halle a. S., 18 (1904), p. 392. 

9 GOETZ, A., and GOULD, W. O.: "The Objective Quantitative Determination 
of the Graininess of Photographic Emulsions," /. Soc. Mot. Pict. Eng., XXIX 
(Nov., 1937), p. 510. 


MR. LESHING: With regard to positive film, our experience from day to day 
shows a considerable and rapid loss of image measured, not in weeks or days, 
but in hours and minutes. The loss of image is sufficient to make us split up our 
night's work into small portions and to avoid keeping our Cinex tests any length 
of time before development. 

MR. COOK: This investigation started where previous investigations have left 
off. The prior work has pointed out that appreciable and important changes in 
the latent image may take place a very short while after exposure. Detectable 
changes may occur even within a few seconds. However, our investigation was 


not primarily concerned with these small time intervals. We found that motion 
picture positive emulsions of various manufacture varied greatly one from 
another in the rate of latent image regression. 

MR. HUSE: It is Hollywood practice to make sensitometric exposures some 
hours ahead of their actual use. For example, the exposures may be made in 
the morning for use that night. Investigation has shown a more marked regres- 
sion of the latent image during the first few hours after exposure beyond which 
the change is slight and progresses very slowly. 

MR. CRABTREE : What is the difference between the results obtained with film 
on a nitrate base and film on an acetate base? Also, what is the effect of high 
humidity and high temperature in the result ? 

MR. COOK: We are not prepared to answer Mr. Crabtree's questions at the 
present time. However, this investigation is being continued and we believe that 
Mr. Crabtree's queries will be answered when pur later findings are published. 


Summary. It would be desirable to have negative exposure control on the basis of 
an exact science. Toward this end the functioning of the eye as it views a subject and 
the photographic reproduction of the subject is studied. The brightness of the sub- 
ject is broken down into its components of reflectance (a constant) and incident il- 
lumination (a variable). The eye compensates for changes in the illumination. 
The "tone" of the object is based on its reflectance. It is this that determines the print 
density used to portray the object. Between the subject's fixed reflectance and the 
print's fixed density lies the variable of negative density. 

A system is proposed whereby given reflectance in the subject is represented by fixed 
density in the negative. Operation of the system involves negative exposure control by 
measurement of incident light. Measurement of effective incident illumination is 
accomplished by a photoelectric meter specifically designed to respond to the three- 
dimensional characteristics of incident illumination. The system is free from many 
of the influences which tend to cause undesirable variations and errors in negative 
exposure. It provides a means of putting negative exposure control on the basis of an 
exact science. 

Every time his camera shutter clicks, probably every photog- 
rapher utters a little inward prayer to the effect that, "Please may 
the exposure be right!" For once the shutter has clicked, the die is 
cast, the Rubicon has been crossed, the deed has been done. There- 
after all the king's horses and all the king's men can not appreciably 
improve the effect of that exposure. The quality, not only of the 
negative, but also of the print to be made therefrom, depends in large 
measure upon the correctness of that exposure. Thus it will be 
appreciated that negative exposure control is a subject of consider- 
able importance to every photographer. 

If negative exposure control had the status of an exact science, 
the photographer would be able to take for granted that his ex- 
posures would be correct every time. As a step toward this very 
desirable objective, the purpose of this paper is to discuss various 
phases of negative exposure control. It is hoped to clarify to some 

* Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received October 1, 

** Hollywood, Calif. 

<> The Society is not responsible for statements by authors $ 

390 D. NORWOOD [j. s. M. p. E. 

extent a subject that has developed some rather confusing phases. 
There will be described some simple tests, in which the action of 
exposure meters will be examined. 

Negative exposure bears a rather close relationship to several 
other elements concerned in picture making. A close examination 
of some of these elements, and the relationships between them will 
serve to bring out some rather interesting facts. 

One of the elements is the appearance of the photographic subject 
to the human eye. It is this appearance that we hope to reproduce, 
within photographic limitations, in the print. Another element is 
the appearance of the photographic subject in the print. Still 
another is the exposure that must be used to make the print from the 
negative, and yet another is the matter of subject brightness and its 
components of incident light and reflectance. 

FIG. 1. Two-card test panel. 

Let us consider a simple case, and note how the various elements 
and relationships are affected. Fig. 1 is a panel composed of cards 
showing two shades of gray. Suppose that a picture were to be made 
to include this panel and nothing else. Any experienced photog- 
rapher, upon viewing the panel, can form a mental picture of how 
a paper print should reproduce what he sees here. His mental 
picture will establish fixed densities of silver deposit on the print, 
which will serve to reproduce what the eye sees first hand. 

Upon turning on additional light, the intensity of illumination on 
the panel, which previously was just nominal, will be increased many- 
fold. During a short transition period the panel will appear intensely 
bright, but after a short lapse of time one becomes accustomed to the 
new level of illumination. 


Now suppose that another picture of the panel is to be made. The 
photographer is asked to visualize the appearance of the paper 
print. In all probability his visualization of the second print will be 
identical with the first. This is not hard to understand when the 
mechanism of the eye is considered. The eye has the property of 
adaptation. It automatically adapts itself to compensate for changes 
in brightness. 

Brightness of any object is a product of two elements: namely, 
incident illumination, a variable quantity; and reflection, a fixed 
quantity. The eye acts to compensate for changes in the variable, 
the incident illumination; the mind judges the appearance of the 
object in relation to other objects by the fixed element, the reflection 

The reflection factors are the means by which we distinguish 
between the two grays on the panel. One gray has a low reflection 
factor, one has a higher reflection factor. It is an appreciation of 
these reflection factors that enables the photographer to determine 
how he will make these grays appear in a print. Thus we can say 
that a fixed relationship exists between the reflection factor of an 
object, and the density of the silver deposit that will represent it in a 

What is the range of reflection factors likely to be encountered 
in photographic practice? Diffuse reflection factors might extend 
from to 100 per cent. However, we find that all objects having a 
diffuse reflectance of 4 per cent or less are called black. Black 
velvet has a reflectance of 2 per cent. At the other end of the 
scale, about the brightest diffused white encountered is white velvet, 
which has a reflectance of 80 per cent. Thus the limits encountered 
in practical work are 2 per cent and 80 per cent, or a ratio of 1 to 40. 

Now, considering negative materials, modern negative emulsions 
have a range of about 1 to 125. This is quite a large range into which 
to fit a subject brightness range of 1 to 40. In fact, it may leave con- 
siderable unused latitude at each end of the negative scale. This 
extended negative latitude has been useful in the past when the only 
means of exposure determination were somewhat lacking in precision. 
However, due to this same extended negative latitude, the develop- 
ment of a precise system of negative exposure control has probably 
been somewhat retarded. The prevailing thought has possibly 
been that if errors in negative exposure could be compensated by 
variations in printing exposure, .then what was the difference? 

392 D. NORWOOD [j. s. M. p. E. 

The arrival of natural color film with its considerably narrower ex- 
posure latitude has changed this attitude in some cases. 

Let us examine the situation developed in this discussion thus far. 
The reflection factor of an object is a constant. The print density 
used to represent this object is a constant. Between these two is 
an intermediate step, which involves the negative density used to 
represent the object. Heretofore this negative density has been a 
somewhat uncertain variable, as can be illustrated by some examples. 
Suppose we are going to make a picture of our panel, and for exposure 
guidance we shall try a photoelectric exposure meter of type re- 
sponding to reflected light. One method of using this meter consists 
in measuring the brightest portion of the subject, also the darkest 
portion, and then computing the mean of the two on a geometric 
scale. This gives a figure representing the mean between the high- 

FIG. 2. Three-card test panel. 

est and lowest values of reflected light. A seating for the camera 
exposure controls is then computed from this figure. 

Now suppose, still using this method, that an additional card is 
introduced into the picture area, as shown in Fig. 2. This card has 
a considerably higher reflection factor than the two previously in the 
picture area. More light is reflected from it. Now, one meter 
reading will be taken on it, and one on the low-reflectance card. 
The mean of the two readings will be quite different from that pre- 
viously obtained. Consequently, the exposure on the film will be 
different from the previous exposure, and the two grays on the origi- 
nal panel will be represented by changed densities in the negative. 
A still different shift in negative density will occur if the two higher- 
reflectance cards were eliminated, and one reading taken on the low- 
reflectance card only. 

Let us examine the action of this type of meter a little further. 
Suppose we take a reading on the low-reflectance card only, and 


make an exposure of this card only, in accordance with the reading. 
Next we take a reading on the medium-reflectance card only, and 
make an exposure of this card in accordance with its reading. Last, 
we take a reading on the high-reflectance card only, and make an 
exposure of this card in accordance with its reading. 

Under conditions of constant illumination the three readings will 
be in proportion to the reflection factors of the three cards. The 
exposure control settings dictated by the meter will vary in inverse 
ratio to the reflectances of the cards, the end-result being that when 
the films are developed the three negatives will all be of the same 
density. The reflectance identities of the subjects will have been 
lost in these identical negatives. 

Now suppose that paper prints are to be made from these nega- 
tives. One print must be of low reflectance to represent its original 
correctly. The next must be of medium reflectance, and the third 
high reflectance. To arrive at these effects in the prints, from three 
negatives of identical density, it is obvious that the first must receive 
a comparatively great printing exposure, the second an intermediate 
exposure, and the third a very small printing exposure. 

This analysis shows that the reflected-light type of meter, even 
when most carefully used, will not give negatives of uniform char- 
acteristics, so that all may be printed with the same printing exposure. 
The property of those meters that causes the above-described result 
also causes exposure errors on short-scale reversal film such as 

A different method of using this type of meter is known as the 
average-brightness method. With this method the meter is held at 
some position where it encompasses the whole scene at once. The 
reading so obtained represents the average brightness of the whole 
scene, and is frequently used to determine exposure control settings. 

In illustrating this case let our scene be limited by a frame which 
includes portions of the two grays on the panel. It will readily be 
apparent that the meter readings will change as the frame is moved 
around to include different relative proportions of the two grays. In 
position A, which includes equal portions of both grays, one reading 
will be obtained. In position B, which includes a large portion of 
high-reflectance material and a small portion of low-reflectance ma- 
terial, a much higher reading will be obtained. In position C, 
which includes a small portion of high-reflectance material and a large 
portion of low-reflectance material, a much smaller reading will be 

394 D. NORWOOD [j. s. M. p. E. 

obtained. This variety of readings for two subjects of constant 
reflectance, under constant illumination, will not give us the uni- 
formity of results we are seeking. 

The author has developed an instrument that circumvents the 
difficulties just mentioned, but before this instrument is described 
a brief discussion of subject and image brightness is needed. In 
looking at the panel of Fig. 3 there is a temptation to classify the 
two grays by saying that one is a dark gray and one is a bright gray. 

FIG. 3 . Varying areas affect the total light 
reflected to the brightness meter. 

This manner of speaking and thinking is inaccurate, and possibly 
leads to a confused conclusion, because if the incident illumination is 
raised, the present dark gray may become many times brighter than 
the former bright gray. A much more orderly way of identifying 
these grays is by reference to their respective reflection factors, 
which are fixed quantities. If the exact figures are not known it 
would be proper to say that one gray has high reflectance and one 
low reflectance. 

These same relative reflectance values will be achieved in the 
prints in order to reproduce as nearly as possible the appearance of 
the original. It will be realized that the print may be viewed under 

April, 1941] 



high or low levels of illumination, just as the original was viewed 
under either condition. The compensating mechanism of the eye 
will be at work in either case to keep the retinal image at the level 
desired by the mind. 

Since the reflection factor of the subject, a fixed quantity, is of such 
importance in identifying the subject, the author proposes a system 
of negative exposure control in which each reflection factor is repre- 
sented on the negative by a given density of silver deposit. Thus 
the identity of any given subject reflectance will be preserved in the 

This idea may be graphically portrayed by means of an H&D 
curve, as shown in Fig. 4. An interesting point to note is that flesh 



FIG. 4. Subject reflectance on D-log E curve. 

tones, having a reflectance between 30 and 40 per cent, fall on a very 
desirable portion of the curve. This system for controlling negative 
exposure is based on the fact that subject brightness can be broken 
down into its components of reflection factor (the fixed quantity) 
and incident illumination (the variable). 

The reflection factor is to be represented in the negative by a given 
density of silver deposit. It then follows that the variable element 
of subject brightness, incident illumination, must be compensated by 
adjusting the lens aperture and the shutter speed. The three 
variables of incident illumination, the lens aperture, and the shutter 
time, are firmly bonded together. A change in incident illumination 
imperatively requires a compensating change in //value or shutter 
time. The scientific determination of the //value and shutter speed 

396 D. NORWOOD [j. s. M. p. E. 

thus inevitably requires a measurement of incident illumination 
rather than reflected light. 

Under the conditions we have set up for this panel (Fig. 3) a 
measurement of incident illumination would be comparatively 
simple. From the position of the subject, or panel, we could point 
a photoelectric meter, properly calibrated, at the light-source, which 
is presumably located beside the camera. The value of the incident 
illumination would be determined by the meter, and from this value 
the proper exposure would be computed. Changes in the panel, 
such as the addition of other cards having different reflection factors, 
would have no effect upon the result. The camera exposure controls 
^^^^^^^^^^^^^^^^^ would always be set to com- 
pensate for changes ki the in- 
cident illumination. Objects in 
jlU the scene would always be re- 

jl corded with given densities which 


f would correspond with their re- 

spective reflectances. 
Three forms of departure from 
this simple case must be recog- 
nized. One of these is due to 
the fact that most photographic 
subjects are not plane surfaces 
like the panel. They are three- 
dimensional, having planes and 

FIG. 5. Modified reflectance on spheri- curves arranged spatially at many 
cal surface. 

different angles. For example, 

consider the sphere shown in Fig. 5. Even if the illumination 
on the sphere were perfectly uniform from the front, the portion 
directly toward the camera would reflect somewhat more light 
toward the camera than the other sides in accordance with Lam- 
bert's cosine law. Due to this law the camera would see the central 
portion as brighter than the sides. This effect, however, is very 
fortunate, since it duplicates the effect that the subject makes upon 
the human eye. The end-result will be that the print will represent 
the sphere just as the eye sees it. 

However, due to this effect it would be desirable to use the term 
"modified reflectance" when considering the portions of a photo- 
graphic subject which are subjected to its influence. This refers 
to reflectance as seen and evaluated from the camera position. So 


much for the effects of surface angularity acting to produce modified 
reflectance values. 

Another type of modified reflectance is produced by local variations 
in illumination within the scene. For example, visualize a white 
stucco garden wall. The sun is shining on it; however, there are 
local obstructions to the primary light, in the form of trees which cast 
leaf shadows on the wall. Under such a set-up we shall find that con- 
ditions of modified reflectance exist. Such a set-up may fall into any 
one of three classifications: 

First, the wall may be almost clear, with only a few leaf shadows. 
In this case the prevailing illumination would be the direct light 
from the sun. The few shadow parts of the wall would be con- 
sidered as having modified reflectance. 

FIG. 6. Balanced side lighting. 

Second, the wall might be almost covered with shadows, with only 
a few shafts of sunlight shining through to strike it. In this case 
the prevailing illumination would be that existing in the shadows. 
The few bright spots would then be the ones having the modified 

Third, the proportions of light and shadow on the wall might be 
roughly equal. In this case the prevailing illumination would be 
considered to be a mean between the direct sunlight and the shadow 
light. The result would be that both highlights and shadows would 
have modified reflectance values with respect to the prevailing 

The concept of "prevailing illumination" on a photographic 
scene is an important one. When the eye looks at a scene the adap- 
tation mechanism functions and assumes some set adjustment. 
This adjustment recognizes the value of prevailing illumination, 



[J. s. M. p. E. 

and compensates therefor. A corresponding adjustment of the 
camera exposure controls is necessary if the negative is to be exposed 
properly. The proper adjustment for these controls may be deter- 
mined from a measurement of the prevailing illumination. As the 
eye adapts itself to view the scenes, so will the camera exposure con- 
trols be adapted. The proposed system thus follows the natural ac- 
tion of the eye ; what the eye sees in the way of tones the camera will 
also see. 

FIG. 7. Prevailing-illumination meter. FIG. 8. Effect of front light on 


The other major form of departure from the simple case comes from 
the fact that illumination on a photographic subject does not always, 
or even usually, come from a point adjacent to the camera. The 
illumination may come from almost any direction with respect to 
the camera-subject axis, and may come from several directions at 
once. This situation makes difficult the use of a plane-surface photo- 
electric cell to measure the incident light, such as was used to mea- 
sure the light on the panel. 

For instance, light striking on the side of a three-dimensional 
subject is quite effective in illuminating it, as illustrated in Fig. 6. 
The plane-surface cell does not properly evaluate this illumination. 

April, 1941] 



In order to get a device which will properly evaluate this illumination 
it is necessary to go to a photoelectric meter which has a three- 
dimensional light pick-up surface. 

Consider a hemispherical surface acting in this capacity as shown 
in Fig. 7. Let it be oriented so that the axis of rotation of the 
hemisphere is on a line with the camera lens, the hemisphere being 
located at the position of the subject and directed toward the camera 
lens. A hemisphere is used because it represents as much of a sphere 
as a camera can see at one time, 
and its surfaces are arranged at 
all the possible angles that could 
be presented by a photographic 
subject to the camera. All illu- 
mination that would be photo- 
graphically effective likewise 
would fall on the hemisphere. 
The hemisphere would receive 
and integrate all such illumina- 
tion at its photographcally effec- 
tive value. 

For example, suppose that a 
single source of illumination were 
located adjacent to the camera. 
This is the most efficient location 
for an illumination source, al- 
though it may be far from being 
the most artistic as to results. 
However, the light from this 
source goes directly to the sub- 
ject and is reflected directly back with maximum efficiency. Other 
factors being equal, such a set-up would call for minimum expo- 
sure setting control. 

Under these circumstances the meter with the hemispherical light 
pick-up, used as described, would receive illumination over the entire 
surface of the hemisphere (Fig. 8). This would result in a large 
needle deflection of the meter, and would dictate a minimum ex- 
posure control setting, which is exactly what the circumstances 

Now suppose that the light-source were moved around through a 
90-degree arc, so that its light would now strike the subject from the 

FIG . 9 . Effect of side 1 ight on c oil ector . 

400 D. NORWOOD [j. s. M. p. E. 

side. Photographers have for many years had a rule of thumb to 
increase exposure controls by one //stop for side lighting, which is 
the condition existing here. 

The side light directly illuminates just about one-half of the 
hemisphere (Fig. 9). With only one-half of the hemisphere illumi- 
nated, the meter reading will be one-half that obtained in the previous 
case. Such a meter reading would indicate twice as great a setting 
for the exposure controls, which corresponds exactly with photog- 
raphers' experience. 

Similarly for all other angles of light approach to the subject, 
the hemisphere will pick up the portion of the illumination that will 
be effective upon the subject, and the meter will evaluate this il- 
lumination. Here then is the ideal instrument for measuring the 
photographic effectiveness of illumination incident on a photo- 
graphic subject. 

The device is inherently free from the influence of a number of 
factors that heretofore were at times sources of error. It is unaffected 
by relative proportions of high and low-reflectance areas in the sub- 
ject, as demonstrated with the panels. It is unaffected by contrast 
of subject, effect of haze, distance of subject from camera, chromatic 
composition of subject, because it never "looks" at the subject, and is 
therefore independent of subject variations. Furthermore, it re- 
quires only one reading per scene for the great majority of all set-ups. 
It approaches the ideal of true instrumentation in that most of the 
factors that invite error due to dependence upon human judgment are 
eliminated from consideration. 

The meter is universal in application, being equally effective for 
interior work and exteriors; and incandescent lights, arc lights, and 
daylight. It is suitable for both motion picture and still photog- 
raphy, and is excellent for both black-and-white film and natural- 
color film. 

In the case of outdoor photography, where large distances may be 
involved but where the lighting is comparatively uniform over the 
area, it is not necessary to use the meter strictly at the subject's 
position. It may be used at any position where the lighting is com- 
parable to that on the subject right beside the camera if desired. 
The axis of the meter, however, should always be arranged parallel 
to the axis of the camera lens. 

A prime feature of the proposed system of negative exposure con- 
trol is that it corresponds to the natural action of the eye in adapta- 


tion to the prevailing illumination on any scene. This feature makes 
it of great value in practical use, since picture making is based upon 
the manner in which the eye first sees the original subject and later 
the printed reproduction of the subject. Another important feature 
is that with negative exposures under precise control by this method 
it is possible to accomplish all normal printing within a very narrow 
range of printing exposure. 

It is realized that there are several factors other than the proper 
measurement of light involved in negative exposure control. Some 
of these are variations in negative processing, variations in negative 
sensitivity and variations in the transmission of different lenses for 
a given lens aperture. In various quarters, however, attention is 
being paid to these variables and their effects are steadily being 

In no phase of negative exposure control does general practice 
seem to be farther from scientific soundness than in that of measuring 
light. It is believed that the use of the system and device described 
here will do much toward establishing a more rational basis for 
negative exposure control, with consequent benefit to all concerned 
in the making of pictures. 


LUCKIESH, M., AND Moss, F. K. : "The Science of Seeing," Williams & Wilkins 
Co., Baltimore, Md. (1931). 

LANGE: "Photo-Elements." 

GOODWIN, W. N.: "The Photronic Photographic Exposure Meter," J. Soc. 
Mot. Pict. Eng., XX (Feb., 1933), p. 95. 

NORWOOD, D. : "Studying Photoelectric Exposure Metering," Amer. Cinemat. 
(Nov., 1939; Jan., Feb., March, 1940). . 

"Photometry," Encycl. Brit. (14th ed.). 


MR. NORWOOD: This meter is calibrated with an arbitrary scale which ties 
in with the Weston system of photo units. Full sunlight gives a reading of 450 
on the scale. The meter is triple-range, in multiples of 10, the ranges being 0-10, 
0-100, and 0-1000 units. It is sensitive down to a value of 0.05 unit. In terms of 
ft-candles, the range is 1.25 to 25,000. 

To compare the range of this meter properly with the range of a brightness 
meter it is necessary to compute the equivalent values in candles per sq-f t encoun- 
tered under given conditions of subject reflectance. 

The reflected-light meter, which measures average brightness of the subject, 
would, if the subject were assumed to have an average reflectance of 13 per cent, 
require a range of 0.05 to 1040 candles per sq-ft in order to be equivalent in sensi- 


Now consider the brightness meter which is used to measure both highlights 
and then dark spots, after which a mean is computed. Assume a reflectance of 2 
per cent for dark spots and 80 per cent for highlights. The brightness meter 
would then require a range of 0.008 to 6370 candles per sq-ft in order to be equiva- 
lent in sensitivity to the illumination meter demonstrated here. 

MR. SIMMONS: How is the sphere constructed? 

MR. NORWOOD : It is made of a translucent, diffusing material, and is hollow. 
Behind it is the photoelectric element. Light striking the sphere is transmitted 
through to the element. 

MR. SIMMONS: Would it be more feasible to build a spherical photoelectric 

MR. NORWOOD: Some sort of protective surface would be required. If glass 
were used it would probably reflect a good deal of the light. 


Summary. The California Consumers Corporation of Los Angeles set aside 
one of its large ice-storage buildings to introduce to the studios a new method of making 
realistic snow scenes. The purpose of the ice-storage building was to furnish a low- 
temperature sound-stage, where water ice could be used for snow, and enable the cast's, 
breaths to become visible, as actually occurs in cold or wintry climates. 

Snow is manufactured on the low-temperature sound-stage by means of specially 
constructed portable blowers, grinding 50-pound blocks of ice and expelling through 
suitable nozzle a fine, aerated snow, directed to the set where and when needed. 

The introduction of Technicolor to the low-temperature sound-stage created many 
new problems in ventilation, due to the low temperature of the atmosphere and quantity 
of air movement needed to remove gases and smoke from the stage during shooting 

The unusual heat load requirements necessitated the construction of external bunker 
systems to augment the existing refrigeration for color production. This was accom- 
plished by the combined use of water ice and ammonia refrigeration in these bunkers, 
giving a total refrigerating capacity of approximately 650 tons in the system to chill 
64,000 cfm of fresh air to 20 F. 

Cold, wintry scenes have necessarily been reproduced through the 
use of artificial means. The Arctic appearances and wintry effects 
have been improved by the addition of visible breath in order to give 
a needed reality to the snow scenes through the use of actual low- 
temperature settings of cold, frozen settings, ice, and realistic snow, 
giving a most stimulating effect to the members of the cast. 

Snow used on the refrigerated stage is manufactured on the set 
through portable snow machines, taking fifty-pound blocks of ice 
and reducing them through a primary crusher, feeding this fine, 
crushed ice to the hopper of a special aerating blower, which, in turn, 
forces the aerated and pulverized ice through an extended flexible 
nozzle, forming a fine, light snow that is placeable in any quantity and 
amount as needed. These portable snow machines are not confined 

* Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received August 
27, 1940. 

** R. Van Slyker Enterprises, Los Angeles, Calif. 


0- The Society is not responsible for statements by authors & 

404 R. VAN SLYKER [j. s. M. p. E. 

to the ice-house but on a great many occasions have been used on the 
Studio's own sound-stage or lots where the snow machine easily 
keeps up with the nominal ice meltage. The photographic value of 
this manufactured snow gives a natural and satisfactory effect, as 
the crystalline luster of nature's snow is captured by the camera. 

The California Consumers Corporation of Los Angeles set aside 
one of its large ice-storage buildings in an effort to introduce to the 
Studios this new method in making realistic ice scenes. The con- 
version of the refrigerated ice-house into a usable sound-stage was 
easily accomplished. In structure the ice-house has a free area of 
140 by 90 ft by 42 ft high, creating no unusual or acoustical problems, 
as the walls of the building were already insulated with 14 inches of 
shavings (for ice-storage requirements), giving the usually acceptable 
noise level of 30 db, and the main problems to be overcome were 
those of heat vs. cold. 

In determining the effect of temperatures for the operation of the 
ice-house without causing noticeable bodily discomfort to the cast 
and technical department, experimental findings show that a range of 
26 to 28F was sufficient to hold any amount of snow and ice on the 
sets and was not severe enough to cause discomfort except in isolated 
cases, allowing a margin of 4 to 6 rise before approximating the 
melting point of the ice. Common colds and catarrhal discharges 
were less prevalent in unit groups than found elsewhere during the 
cast's stay in the refrigerated stage. The bodily temperature changes 
were modified through the installation of a vestibule maintained at an 
intermediate temperature in order to avoid abrupt temperature 
changes. The serving of ample heated liquids greatly reduced any dis- 
comforts the low temperature may have caused to members of the 

During the operation of this stage, operating factors governing the 
refrigerated stage are entirely unlike those of any other sound-stage, 
due to the extreme low operating temperature required to preserve 
the ice and snow settings and to maintain the set temperature so as to 
gain the visibility of the cast's breath. Among other operating 
factors is the needed 25 to 30 tons of refrigeration required to offset 
the heat transfer of the walls. 

The heat load based on lamp amperage for black-and-white 
photography ranges from 1500 to 4000 amperes, necessitating a 
minimum of 185 tons of refrigeration to offset this particular heat 
load, which may be readily calculated to the following formula : 


3.42 BTU = 1 Watt X 1000 W. - 5619 BTU 
Amperes X BTU = Heat Load -f- 333,000 

Heat Load X 3420 BTU 

12,000 M.P.P.T. - Load Inter, of Ref. 

One serious factor in the operation of the refrigerated stage is that 
of replacing the exhausted, vitiated, and other foul air from the room 
with chilled fresh air, in order to prevent the warming of the prevail- 
ing interior temperature through the introduction of warm, outside 
air. This was handled by the installation of a large air-precooling 
bunker 'system consisting of approximately 6000 feet of 2-inch am- 
monia pipe, in addition to the existing refrigeration system of nearly 
4 miles of ammonia piping installed on the ceiling of the stage. 

The bunker system equipment consists of a 25,000-cfm positive 
pressure blower passing the warm room air, together with approxi- 
mately 30 per cent of fresh, outside air, through the coils, and ex- 
hausting the cooler air into the stage at approximately 1 inch of 
pressure, the foul air in turn being expelled through a series of suitable, 
controllable exhaust hatches located in the stage ceiling some 42 feet 
from the floor. 

After tests were made of this initial air flow during actual opera- 
tion, it was decided to increase the air flow over the ceiling coils by 
the installation of four utility air-screw fans recirculating the warmer 
air near the ceiling, giving an additional air circulation of some 16,000 
cfm of free air across the main ceiling refrigerating coils. 

The misty vapor accompanying ice or snow scenes is greatly elimi- 
nated on the stage by virtue of this large, cold, dry air flow. The 
vapor or fog is in reality water vapor released by the ice when the di- 
rected rays of the lamps strike it. After flowing the damp, vapor- 
laden air over the cooling coils, it precipitates a large percentage of 
the air-borne moisture. The slight vapor released by the ice is thus 
absorbed and a consequent reduction of the total vapor in the stage 
atmosphere occurs. 

With the introduction of Technicolor photography to the ice-house, 
the large increase in lighting requirements over black-and-white 
photography caused many problems in adapting the refrigerated 
stage to handle the increased heat and gas content of the air gen- 
erated by the numerous carbon arc lights required for the Technicolor 
process. Heretofore recirculation of 70 per cent of the interior air 
was possible, due to the prevalence of incandescent lighting used in 

406 R. VAN SLYKER [j. s. M. p. E. 

black-and-white photography. This recirculation could not be con- 
tinued due to the high gaseous content of the air through the use of 
nearly all arc lights. The increased gaseous volume of air required 
the addition of approximately 48,000 cfm of chilled air in order 
properly to ventilate the stage and maintain the correct dry and wet 
bulb temperature conditions. 

In determining a method of ventilating the stage during Techni- 
color.shooting, it was decided to enter air at approximately 20F at 
floor level and allow the heated air to rise and escape through the 
ceiling exhaust hatches. During the calculation of the required re- 
frigeration tonnage to maintain the stage floor level at approxi- 
mately 20F during the filming of Technicolor productions, it was 
found that the refrigeration plant equipment supplying the stage 
could handle an additional heat load of only 25 per cent, as during the 
summer season of the year the remainder of the plant refrigeration 
equipment was needed for ice manufacture and cold-storage facilities. 

The problem was additionally complicated due to the present cold- 
storage plant set-up, as the various storage departments were on 
different suction pressures; also due to the unusually high gas con- 
tent of the stage atmosphere it was essential that only clean, fresh 
air should be used on the stage. 

Again, due to the short time available for equipment installation, 
it was not deemed feasible to try to install additional compressor 
capacity, or to couple a brine-cooling system to the ice-manufacturing 
section of the plant. Further complications were those of obtaining 
proper air circulation on the stage due to the type of construction 
and location of the set. Adequate air circulation was further hindered 
by a large cyclorama hanging from the walls of the stage. 

The total connected electrical load for this Technicolor production 
was approximately 17,000 amperes, which alone would have required 
540 tons of refrigeration on the stage had this connected load been 
used in its entirety. 

Fortunately, the total connected lamp load of a set is almost never 
used at one time, and the maximum recorded load on this production 
was approximately 9600 amperes. Again, it was noted that the aver- 
age burning period of the lamps in a black-and-white production is 
approximately 20 to 30 minutes out of each hour. With Technicolor 
the general average is increased to 30 or 40 minutes in each hour, and 
we were able to eliminate approximately 50 per cent of this heat load 
by placing many of the lamp resistances in the attic immediately over 


the stage sets, permitting the attendant resistance cables to exit 
through the ceiling exhaust hatches. 

By taking every possible advantage, we were able to reduce the 
total available heat load of the lamps to approximately 140 tons of 
refrigeration. While this figure may seem low, it must be remembered 
that the resistances of the lamps account for approximately 50 per 
cent of the total ampere heat load input. Again, in order to obtain 
the required air change in this building, and from the experience of 
the Studios on stages of identical size, it would require a total of 
64,000 cfm of chilled fresh air to do so. This excessively high volume 
of fresh air in itself requires a refrigeration capacity of over 500 tons 
to reduce the normal wet-bulb temperature of approximately 70 
to 20F. The existing temperature on the exterior of the building 
during this particular time was approximately 83 to 90F (in the 

The cold air supplies were divided so as to be handled by three 
separate blower systems, and these were of slightly different sizes in 
order to achieve a desirable flexibility. 

The installation of two blowers of 18,000 cfm each and one blower 
of 28,000 cfm gave the required total volume of 64,000 cfm. The re- 
frigeration load was divided into two divisions to make use of the 
available refrigeration plant capacity of 200 tons of refrigeration and 
to apply the required refrigeration difference through the use of wet 
ice bunkers. 

The division of the load in this manner worked out satisfactorily, 
as it was possible through the use of wet ice bunkers to reduce the tem- 
perature of the incoming air to approximately 40F dry-bulb and 
39F wet-bulb, as had been proved possible by previous experimenta- 
tion. This would account for nearly 450 tons of the 650 tons' total air 
load refrigeration so that a final 200 tons were left for ammonia re- 
frigeration. This permitted the 200 tons' available capacity of the 
refrigeration plant to reduce the air to the required floor temperature 
of 20F through the use of a series of finned coils. Having settled 
the disposition of refrigeration loadings, construction was started on 
three separate bunkers to house the blowers and cooling equipment, 
including in these bunkers a capacity of 450 tons of wet ice for the 
three primary stages. The hourly ice usage was 450, plus twenty-four, 
equalling 18.75 tons of melt per hour. 

Dividing this ice consumption between the three ice bunkers accord- 
ing to air volume : 

408 R. VAN SLYKER [j. s. M. P. E. 

Bunker A to be connected with existing 50-ton straight pipe coil is as follows : 

18,000 cfm bunker 

18.75 tons of ice total X nnn , - = 5.3 tons of ice 
per hour 64,000 cfm - total pef hour 

Bunker B to be connected to 28,000-cf m fan and center of building : 

28,000 cfm bunker 

18.75 tons of ice total X - . -- . , - = 8.2 tons of ice 
per hour 64,000 cfm - total pr ^ 

Bunker C to be connected to 18,000-cfm fan and end of building: 

18,000 cfm bunker 

18.75 tons of ice total X - = 5.3 tons of ice 
per hour 64,000 cfm - total pef 

The three ice bunkers were set up as vertical chambers, which 
were loaded from the top with ice broken in about 25-pound pieces, 
and through which the outside air was drawn vertically from top to 
bottom. This arrangement allowed an even spread of ice across the 
area where the air contacted the ice, and permitted no voids to occur 
as would have happened if air had been drawn through horizontally 
and the ice had melted down to the lower side of the cooling area. 
Also this arrangement allowed the loading to be done through the 
same opening through which the air entered, so that loading could be 
done while the system was in operation. The ice bunkers then con- 
sisted of vertical shafts, with open tops, placed alongside the evapora- 
tor chambers and opening along the bottom of the adjoining side, so 
as to allow the air to leave the bottom of the adjoining side and the 
bottom of the shaft, and enter the evaporator chamber horizontally. 

Supporting the column of ice was a grille set at the level of the top 
of the evaporator plenum. The grille at this level permitted the air 
to be drawn completely through the ice while travelling in the same 
direction, so that the ice was melted evenly and pockets were avoided. 
The required height and area of the shaft were based upon (1) the 
average density of the ice, in 25-pound blocks; (2) the optimum ve- 
locity of air through ice for transfer of cooling, but without excessive 
resistance and possible noise; (3) the necessary height of ice column 
to produce the required temperature reduction, but without excessive 

Through previous experiment, it was found that one cubic-foot of 
the scored ice as broken into 25-pound blocks averaged about 27 
pounds. It was also found that the effective open area for air travel 


through the shaft was approximately one-half of the total area, so 
that the effective velocity through the ice was approximately twice 
the velocity of the shaft when empty. Also, it had been proved in an 
experimental tower that 8 feet of ice column were necessary to reduce 
ah- at 80 to 40. Based on these factors, the average dimensions of 
these three bunkers were 8X8 feet in area and 16 feet in height. 

The coils for B bunker were four four-row sections 72 inches X 
24 inches with three fins per inch, to give a tonnage of 92, with a ve- 
locity of 683 fpm. The coils for C bunker were two four-row sections 
72 inches X 30 inches with three fins per inch, to give a tonnage of 58, 
with a velocity of 600 fpm. The nozzles for the brine spray system 
to keep the B coil defrosted were Binks No. 11 Rotojets, with a 1 /2- 
inch pipe connection and 3 / 8 -inch orifice, and with a capacity of 4 
gpm at a pressure of 10 pounds per square-inch. Twenty of these 
were used, arranged four rows wide by five high, and set approximately 
12 inches from the face of the coil. Twelve of these same nozzles set 
four wide by three high were used for C coil. 

The brine pumps used for the defrosting were 80 gpm for the B 
coil and 50 gpm for the C coil, with each providing for a head of 40 

Each ice bunker was constructed to carry sufficient volume of ice 
for approximately a 2V2-hour run so that one loading crew could 
alternately load the bunkers and, allowing 45 minutes for loading, 
could go continuously from bunker to bunker. 

Each bunker had the following capacity : 

Bunker A 2.5 hours X 5.3 tons/hour = 13 tons of ice capacity 
B 2.5 hours X 8.2 tons/hour = 20 tons of ice capacity 
C 2.5 hours X 5.3 tons/hour = 13 tons of ice capacity 

These three ice bunkers were set up in vertical chambers, which 
were loaded from the top with ice broken into about 25-pound pieces. 

The resulting performance of this new refrigeration equipment can 
be seen from the following chart taken during one of the shooting days. 
These readings, which were taken under typical conditions of weather 
and lighting load, were made at 2 : 00 P. M., August 2, 1939. The maxi- 
mum lighting load used that day was 8500 amperes: 

Outside temperature 83 F 

Temperature of air off ice, Bunkers A, B, and C 40 

Temperature of air off coil, Bunker A 20 


B 21 

C 19V 2 

Room temperature, before shooting 2:00 p. M. 21 

Room temperature, after shooting 2:30 P. M. 23 

Temperature under light parallels 28 

Temperature on light parallels 28 ft from floor 34 

Temperature in exhaust outlets from attic 74 

Average ice usage 16 tons per hour 

Ammonia suction pressure on finned coils 5 Ib/sq-in. gage 

Ammonia suction temperature on finned coils 17 F 

Average fresh air usage 62,500 cfm 


Summary. A brief discussion of the desirable qualifications for projectionists, 
both as to theoretical and practical training. Reference is made to a system of 
licensing and examination in the Province of Alberta, Canada, as contrasted -with 
unsatisfactory conditions prevailing in a number of parts of the United States. 

Some will doubtless regard what is said here as merely a re-discus- 
sion of an old story, namely, the importance and necessity for the 
further practical and technical education of projectionists in their 
art. Granting that, technical members of our Society and the en- 
gineering departments of equipment manufacturers and producing 
companies must in addition recognize the extent to which their work 
may be nullified by the failure of many theaters to employ projec- 
tionists able to supply adequate evidence of capability in handling 
the equipment and products placed into their charge. 

Such failure on the part of the theaters has worked considerable 
harm. Replacement of projectionists lacking in technical knowledge 
and skill by others well versed in the technique and practice of pro- 
jection would produce great improvement in results and economy of 
equipment and operation. The existing situation seriously discredits 
the work of the scientists and engineers in the eyes of the theater pa- 
trons as well, and the author feels that the Society has not taken 
sufficient cognizance of the fact. This is not a blanket charge against 
all projectionists. There is a constantly increasing number who, 
having sensed the importance of their work, are studying its technic 
and practice to such good purpose that they are able to do full justice 
to the apparatus and productions handled by them. Such men are a 
credit to their profession. On the other hand, there are still some 
men holding positions as theater projectionists who are equipped 
with little if any technical knowledge; who have little understanding 
of the forces they are called upon to handle; and who in many in- 
stances are possessed of mediocre mechanical skill. 

* Quigley Publishing Co., New York, N. Y. 


$ The Society is not responsible for statements by authors $ 

412 F. H. RICHARDSON [j. s. M. p. E. 

Such a state is unfair to the technician, the engineer, the manu- 
facturer, the producer, the exhibitor himself, and to the theater-going 
public, which in the final analysis supports the entire industry. The 
theater patron is unable to locate the source of a fault in projection 
and makes no effort to do so, but it is generally known that he always 
attributes the fault to the equipment or the film. 

A few efforts have been made to rectify the situation, or at least to 
ameliorate it. As an example, we may cite the regulations of the 
Province of Alberta, Canada, recently promulgated not only for the 
purpose of reducing possible fire and other hazards in motion picture 
theaters, but to improve the performance as well. Alberta requires 
one year of service as a licensed apprentice projectionist before a final 
license can be granted, with the exception that projectionists from 
other places should be able to convince the authorities of their com- 
petency and may receive their licenses without serving the appren- 
ticeships. During the period of apprenticeship the applicant must 
have at least 300 hours of actual service in a theater projection room 
under the supervision of a first-class projectionist. During all this 
time he is not permitted to assume charge of projection. Violation 
of this regulation may result in the loss of a license of the licensed 
projectionist supposed to be in charge. During the apprenticeship, 
quarterly examinations are held to ascertain the rate of progress. 
At the end of the year, after 300 or more hours of actual service, the 
apprentice may take his examination. If the result is satisfactory, 
a third-class license will be awarded permitting him to act as projec- 
tionist in a theater containing 350 seats or less. If the theater has 
more than one projector, a second licensed man must be constantly 
on duty. At the end of a year of service under the third-class license 
the man may appear again for examination. If the result is again 
satisfactory, he will be awarded a second-class license permitting him 
to take charge of projection in theaters of 500 seats or less ; and, again, 
if there is a second projector, there must be a second licensed projec- 
tionist to assist him. During this year, 500 hours or more of actual 
projection work is required. At the end of the year of the second- 
class service the man is reexamined, and if found competent will 
finally receive a first-class license. 

It may be observed that Alberta takes its license laws very seri- 
ously. The laws specify, in addition, that certain subjects be in- 
cluded in the examinations for each class of license, and it may 

April, 1941] THEORY vs. PRACTICE 413 

safely be assumed that the holder of an Alberta first-class license is a 
thoroughly competent projectionist. 

All Canadian provinces proceed more or less along the same gen- 
eral ideas, and definite assurance must be provided at each stage that 
actual progress in technical knowledge and experience has been made 
before the next class of license may be granted, and the learning and 
experience must be gained under the tutelage of a first-class licensed 

On the other hand, let us look at the situation in one of the states 
of our own country, which, for obvious reasons, will be unnamed. 
Apparently the state recognizes that some protection is due its citi- 
zens against the employment of incompetent projectionists, but aside 
from the recognition of general safety practices in relation to fire 
hazards in public meeting places, little is required in the way of ac- 
tual competence in projection. Regulations are extremely vague, 
and only a "practical" examination is requred. No attempt is made 
to be sure that even the examiner himself is competent to give an 
examination on theater projection, and it is left entirely to the ex- 
aminer as to what shall constitute a "practical" examination. But 
that is not all. In many cities, towns, and villages of the country, 
there is no protection at all other than by law. In many instances, 
any school boy with no technical knowledge or practical experience 
may take charge of projection in a theater, and not only abuse the 
equipment and film, but jeopardize the audience and give a "black 
eye" to the entire industry. 

The lack of laws specifying anything further than "practical" ex- 
aminations in first-class cities lays the entire system open to political 
and other abuses, wherein it might be possible for examiners to reduce 
the meaning of the term "practical" to very low values at the request 
of political "higher-ups." 

One state at one time made provision for apprenticeship prior to 
examination for the final license, but the law was later thrown out 
because it contained no clause permitting the licensing, without ap- 
prenticeship, of projectionists from other states who could prove 
thorough competency. 

There seems to be wide opposition to the idea of apprenticeship, 
chiefly on the grounds that there may be in some parts of the country 
more projectionists available than positions. This objection might 
be overcome by a requirement that all projectionists now holding 
licenses be reexamined, under a really competent examiner. 


The question arises as to what the Society of Motion Picture Engi- 
neers can do in the matter. It would seem that the Projection 
Practice Sub-Committee, of our Theater Engineering Committee, 
could examine the situation and recommend to the states a set of 
regulations that might be used in examining applicants for projec- 
tionists' licenses. They might also prepare a list of subjects upon 
which such applicants should be examined, and determine the scope 
of knowledge and experience that the applicants should have before 
being permitted licenses. Suitable provisions could be included in 
such recommendations for the apprenticeship which has been pro- 
posed here. 


The history of the sound picture dates back to 1877, the year of 
Edison's announcement of the phonograph, which brought the reali- 
zation that the human voice could be mechanically reproduced. 
Between that time and the time of the successful demonstration of 
the motion picture in 1889 there were a number of scientific dreamers 
who hoped to make talking photographs. They tried to synchronize 
the human voice or music to slide photographs ; and with the advent 
of the "living pictures" these experimenters tried to make of them 
talking motion pictures. 

Wordsworth Donisthorpe was, according to present records, the 
first to suggest and experiment with the idea of talking photographs. 
He wrote a letter to the editor of Nature which was published in the 
issue of January 24, 1878, under the title "Talking Photographs." 
In it he said : 

"By combining the phonograph with the kinesigraph I will undertake not 
only to produce talking pictures of Mr. Gladstone which, with motionless lips 
and unchanged expression, shall positively recite his latest anti-Turkish speech 
in his own voice and tone. Not only this, but the life-size photograph itself 
shall move and gesticulate precisely as he did when making the speech, the words 
and gestures corresponding as in real life." 

He adds further, in effect, that he took his photographs at intervals 
of half or quarter seconds, with an exposure of an eighth second. 
The finished pictures were mounted upon a long paper band. 

His kinesigraph, or camera, was so arranged that as each exposed 
glass plate dropped out of the way, it was replaced by the next plate 
which was directly behind. 

The phrase, "with motionless lips and unchanged expression," 
is evidently a reference to the fact that Donisthorpe recorded only a 

Presented at the 1934 Spring Meeting at Atlantic City, N. J.; revised 
February, 1937; January, 1941. 

** Curator, Motion Picture Division, Los Angeles Museum, Los Angeles, Calif. 

*> The Society is not responsible for statements by authors <> 

416 W. E. THEISEN [j. s. M. P. E. 

series of static poses, which resulted in the loss of facial expression 
and intermediate progressive movement. Because he could photo- 
graph only at the rate of two to four pictures a second, as compared 
with the present twenty-four each second, expression and action 
conforming to the sound record were lacking, as several words would 
be spoken while each still picture was viewed. 

Since the Edison phonograph at that time was far from perfect, 
very little came of Donisthorpe's experiments. One of these early 
Edison phonographs with soft wax records in the possession of the 
author corroborates the phrase, "the wee small voice," by which the 
phonograph was then known. The words are barely distinguishable 
if they be known in advance and the tunes are only approxima- 
tions of the intended tunes. 


The first two men who concerned themselves with photographically 
recording sound were Czmark of Vienna, who, in 1862, photographed 
the vocal cords in action, and Alexander Blake of Brown University, 
who, in 1878, conducted a series of experiments in photographing 
the vibrations of a pencil of light from a mirror attached to a micro- 
phone diaphragm. In Blake's experiments the photographic plate 
was kept in motion by a clockwork mechanism. 

Charles E. Fritts applied for a United States Patent on October 22, 
1880, which covered methods of recording sound photographically. 
In the claims of his patent are specifications for shutters, and various 
optical systems are covered for focusing the shutter or slit records, 
which were made on a long photographic band. 

Fritts suggested that the sound might be reproduced by causing 
a beam of light or other radiant energy passing through the moving 
record to fall upon bars of selenium, which would respond in such a 
way as to regulate an electric current and thus operate a diaphragm. 
He (and many others, subsequently) tried to reproduce sound by 
passing a light beam through a moving photographic sound record, 
thus permitting light of varying intensity to pass to the selenium. 

The brief of the Fritts' patent twenty-six pages long covers 
the basic elements of sound recording as we now know them, and 
shows a deep insight into the problems that have presented themselves 
subsequently. A system of amplification is covered, various types 
of slits and sound-tracks (or "patterns," as he calls them), and many 
other features. This amazing patent was granted thirty-six years 








Patented Oct. 31, 1916. 


/'if/, /Jf. 

FIG. 1. Section of U. S. Patent by C. E. Fritts. 



[J. S. M. P. E. 

after its application, on October 31, 1916, as number 1,203,190* 
(Fig. 1). 

Georges Demeny introduced his chronophotophone, in October, 
1893, although he had demonstrated it earlier. 1 ' 2 It was a device 
that synchronized a photographic record and slide photographs 
which were mounted around the periphery of a glass disk. In the 

FIG. 2. Projection room (about 1894), showing phonograph at- 
tached to kinetoscope for synchronization of sound and picture. 
(Century Magazine, 18, 207, 1894.) 

slides were attempts to portray motion that had been photographed 
in the "Photographic Gun" which Demeny had devised about ten 
years earlier. 3 


In the meantime Edison had approached the problem of giving 
eyes to his phonograph. In The History of the Kinetoscope, Kineto- 

*In this patent are descriptions of a string galvanometer, push-pull recorder, 
narrow slit, rubber roller for damping the recording drum, record wear which 
destroys the high frequencies, and so forth. 



We have at last succeeded in perfectly synchroniiing 
music and moving pictures. The following sc.nes 
are very carefully chosen to fit the words and the 
songs, which have been especially composed for 
these pictures. 

LOVE AND WAR. Uncholeric. 

-The above is an illustrated song telling the story of a hero 
who leaves for the war as a private, is promoted to the rank of 
captain for bravery in service, meets the girl of his choice, who is 
a Red Cross nurse on the field, and finally returns home triumph- 
antly as an officer to the father and mother to wh^m he bade good 
bye as a private. Length 200 feet, complete with words of song 
and music. $15-oo 


B^^A side-splitting subject, showing the mistaken tramp's arrival at 
the Wm. Waldorf Astor mansion and being discovered comfortably 
asleep in bed. by the 'ady of ihe house. Length 100 feet, com- 
plete with words of song and music. . . $20.00 
Without music. . (Unchurch) $15.00 

Complete Set, aa subjects, about 1,950 feet, $400.00 

OPERA OF MARTHA. Unbeflecht. 

The Second Act of this beautiful opera. Consists of five 
scenes, about 1,300 feet in length. I. Duet outside the 
Inn. a. Quartette inside the Inn. 3. Spinning Wheel 
Chorus. 4. Manna singing " Last Rose of Summer." 
5. Good Night Quartette. This film shows a quartette of 
well-known opera singers acting and singing their pans in 
ibis ever popular opera. The subjects are taken with the 
greatest care and the films manufactured by the Edison 
Manufacturing Company. 

Managers can arrange to produce this exhibition through- 
out the country, and can obtain a quartette of church 
singers to remain behind the scenes and sing the pans and 
produce a remarkably fine entertainment, besides giving'a 
local interest to the same by utilizing local talent. If it is 
desired to do so, however, the quartette can be engaged .to 
travel with the exhibition. 

Other operas and plays in preparation 

Complete Sat, 5 scenes, about 1,300 feet, $320.00. 

Special Notice. Thesr films are taken under license / 
7homat A. Kdiso*, whose patents cover moving photo- 
graphic films. 

FIG. 3. Portion of a page in an Edison motion picture 
catalog, dated March, 1900. 

graph, and Kinetophonograph, written by W. K. Laurie Dickson and 
published in 1895, is reproduced in Edison's handwriting: "In the 
year 1887 the idea occurred to me that it was possible to devise an 
instrument which would do for the eye what the phonograph does 
for the ear, and that by a combination of the two all motion and 

420 W. E. THEISEN [J. S. M. P. E. 

sound could be recorded and reproduced simultaneously." That 
briefly sums up Edison's purpose for the motion picture. All his 
early experiments to achieve this purpose were closely patterned after 
the phonograph. 4 

Dickson's history interestingly tells of a studio room constructed 
at the Edison plant at Orange, N. J., for the purpose of creating 
motion pictures, and a sketch of the interior of this studio shows a 
camera and phonograph coupled together for taking a "talkie," 
(Fig. 2). The Kinetophone Edison's sound motion picture device 
was commercially introduced about the same time as his silent 
kinetoscope, or "peep-show," was put into the "peep-show parlors." 
The first of these parlors was opened by the Holland Brothers on 
April 14, 1894, at 1155 Broadway, New York, N. Y. 

Shortly thereafter, many parlors installed kinetophones, in which 
patrons, after ^putting a coin in the slot, could peer through an eye- 
piece at moving pictures and, at the same time, by means of ear-tubes 
somewhat similar to the physician's stethoscope, listen to a sound 
accompaniment. The early catalogues published by Edison listing 
his motion pictures carried a list of orchestral records that had been 
specially recorded for playing with the pictures. 

One of these catalogs of March, 1900, on display at the Los 
Angeles Museum, itemizes several pictures that could be purchased 
with or without sound records (Fig. 3). An explanatory note in 
this catalogue states, "We have at last succeeded in perfectly synchro- 
nizing music and moving pictures. The following scenes are care- 
fully chosen to fit the words and songs, which have been especially 
composed for the picture." The price for a 200-ft. picture, Love and 
War, was $45. 

The kinetophones were not successful, but Edison continued to 
toy with the idea. In 1910 he introduced another system designed 
to show sound pictures in the comparatively large film theaters that 
had, in the meantime, made their appearance. This system con- 
sisted of a phonograph placed upon the stage, run by a long wire 
belt connected to a projector at the rear of the theater. It enjoyed 
no extensive use beyond approximately a four-month run at the 
B. F. Keith Theater in New York. 

The idea of the talking picture was firmly established. It seemed 
logically possible to the first experimenters that the phonograph could 
be synchronized to a projector. Among the first specifications of 
means for achieving synchronization, besides those already men- 


tioned, was that of George W. Brown, who was granted a United 
States patent on February 9, 1897. Two years later, on July 31st, 
a German patent was granted to L. A. Berthon, C. F. Dussaud, and 
G. F. Jaubert, in which were specifications for synchronizing a pro- 
jector and phonograph. 

Valdemar Poulsen patented a novel sound-recording system in 
Germany on April 21, 1900. He passed a steel ribbon between two 
electromagnets energized by a current that was modulated in a 
microphone. When the diaphragm of the microphone was bom- 
barded by sound, the fluctuations were produced in an associated 
electric circuit which in turn induced a magnetic pattern in the steel 

In the Scientific American of July 29, 1901, Ernst Ruhmer an- 
nounced a method of recording a sound-track photographically upon 
motion picture film. It was a fluctuating arc device which he called 
the photographophone. 5i 6 


The British Patent Journals record a number of patents granted 
to Leon Gaumont and H. H. Lake in collaboration during 1901-03, 
dealing with various systems for synchronizing phonographs and 
projectors by means of gears or brushes on the armatures of driving 
motors. The patents specify loud speaker systems, and suggest the 
use of loud speakers behind the screen. The speakers were intended 
to be carried about, following the moving images upon the screen. 

Gaumont's first public demonstration was given on November 7, 
1902, at the Societe Frangaise de Photographic, when he presented, 
among other things, his own talking portrait. The system used 
for this demonstration consisted of a specially designed gramaphone 
from which a projector was operated by means of a flexible shaft. 
Very shortly this flexible shaft was replaced by an electrical con- 
nection. In order to start the phonograph and projector simultane- 
ously, a "current distributor" was attached to the phonograph 
shaft to keep the speed of the two synchronous and uniform. 

In 1903, the system was shown at the Musee Grevan, and a little 
later at the Theatre Du Gymnass, under the name Phonoscenes 
Nothing much was heard of the system until 1910, when a talking 
picture of Professor D'Arsonval was presented to the Academic des 
Sciences on December 27th. It was brought to this country in 1913, 
during which year it was exhibited at the 39th Street Theater, New 

422 W. E. THEISEN [J. S. M. P. E. 

York, on June 5th-7th. In the meantime the Gaumont Film- 
Parlant had been showing nightly at the Gaumont Palace in 
Paris. 7 ' 8 ' 9 ' 10 The Gaumont sound apparatus was known as the 
Chronophone. Subsequently to 1926, Gaumont worked with the 
Petersen-Poulsen sound-film system. 11 


During 1903-04 Oscar Messter was granted several German pat- 
ents on methods of synchronizing phonographs and projectors. 12 
One claim granted to him in an English patent dated October 19, 
1903, specifies a synchronizing mark to serve as a starting guide. 
Messter's system, known as the auxtephone, utilized compressed air 
in the speaker for amplifying the sound. It is said that by means 
of air-compression amplifiers, Messter surmounted one of the chief 
difficulties of the sound picture pioneers, namely, sufficient volume. 
It would seem, however, that Gaumont had sufficient volume in his 
sound pictures, because, in 1913, he is said to have successfully 
shown talking pictures for a time at the Gaumont Palace, previously 
the Hippodrome, to audiences of 4000 persons. 

In 1904, the cameraphone, introduced in New York by James 
Whitman, enjoyed extensive commercial success over its contempo- 
raries. A flexible-shaft arrangement for maintaining synchronism 
between the phonograph and the projector was patented in France 
in 1907 by Georges Pomarede. 13 

Among others to develop and patent phonographic talking picture 
systems in the next few years were W. C. Jeapes, 14 who applied for a 
patent in 1909 on his cinephone, and Cecil Hepworth, 15 who intro- 
duced the vivaphone at about the same time in England. These 
two English experimenters devised an interesting though somewhat 
impracticable method of synchronization. The cinephone, brought 
out commercially by the Warwick Trading Company, had a rotating 
pointer, driven by the gramaphone motor, on the front of the grama- 
phone cabinet, and the progression of the sound record could be 
followed as the pointer turned past a series of green lights on a dial. 
To maintain synchronism of the sound and the picture, the motion 
of the pointer on the gramaphone was made to correspond to that of 
a second pointer photographed on the film. The projectionist con- 
trived to keep the two in step by manually varying the speed of the 



Likewise the vivaphone had a pointer arrangement for synchroni- 
zation. The pointers were actuated by a pawl and two electromag- 
nets, one each for the projector and the gramaphone. The system 
was adaptable to any projector. 16 

A somewhat similar system was devised by Thomassin, with the 
difference that the pointer was held stationary by an electromagnetic 
escapement, and any lack of synchronism was indicated by move- 
ment of the pointer. 17 

E. H. Amet was granted numerous United States patents during 
1912-18 on a system called Audo-Moto -Phone. 1 * These patents 

FIG. 4. E. H. Amet and Mrs. Amet, with their disk sound and picture 
recording equipment 1912. 

covered electrical pick-up methods of recording, and sound devices 
in a variety of forms. That Amet was particularly interested in 
talking pictures may be judged from the fact that the SMPE Histori- 
cal Collection at the Los Angeles Museum includes about thirty-six 
different experimental devices made by him. Newspaper items 
from various California towns where Amet's sound pictures were 
shown were profuse in their praise. During 1914-16, Amet developed 
a "balanced" microphone, which was patented. 19 

For synchronizing, Amet relied upon a connecting shaft between 
the projector or camera and the phonograph (Fig. 4). His "amplifi- 



[J. S. M. p. E. 

cation" through the theater was effected by "loud speakers" set at 
various vantage points. 

In order to give an idea of the many ramifications of the efforts 
made to realize the sound picture, mention must be made of the work 
of Katherina von Madelar, who was granted her first sound patent 
in 1916, on a system of recording a sound-track by means of an electri- 
cally heated needle. 20 The needle 
was attached to a microphone 
diaphragm, which, upon being 
caused to vibrate by the imping- 
ing sound-waves, caused the 
heated needle to cut a waving 
line in a celluloid ribbon. Another 
novel von Madelar system in- 
volved a rotating cutter, which 
turned in contact with the edge 
of the film. The cutter received 
the sound vibrations from the 
microphone diaphragm, and re- 
corded them as a jaggedly cut 
pattern upon the film edge. The 
system was called the projecto- 

Another system of synchroniz- 
ing the cinematograph and the 
gramaphone was devised by Wil- 
liam H. Bristol, and termed the 
Bristolophone. It was patented 
in the United States in 1917, and 
was used to some extent after 
the advent of the talkies. 21 
There were many, many others. An investigation of the patent 
records indicates that the researches undertaken to create the talking 
picture were as diverse as were the experimenters, and that often 
the dreams of the experimenters led them far afield. In England 
alone, for the years 1908-09, twenty-seven patents were granted. 
These many undertakings were not without value, however; the 
practical and the impractical researches both have aided in accumu- 
lating the knowledge that has led to the perfection of the talking 

FIG. 5. Lauste's string recorder; 
used by Lauste about 1910-11 for 
recording sound-on-film. 




In the meantime there "were [parallel developments by other men 
who tried to record the sound on the photographic emulsion. In 
tracing the history of the sound-film, the work of Eugene A. Lauste 
is generally regarded as very significant. Lauste made his first recorder 

FIG. 6. Lauste's string recorder (open), 
showing diamagnetic double vibrating wire 

in 1904, along the lines of earlier apparatus made by both Fritts and 
Blake. It consisted of a box containing a slit, or narrow aperture, 
through which a light-beam was directed by a mirror attached to a 
microphone diaphragm. Movements of the diaphragm caused by 
the sound vibrations produced movements of the mirror and the 
light-beam, the vibrations being recorded upon motion picture film. 
Lauste continued to improve and to elaborate his devices in an 
attempt to record the picture and the sound upon the same film. 



[J. S. M. p. E. 

On August 11, 1906, Lauste, in collaboration with Robert T. Haines, 
an Australian experimenter, and John St. Vincent Pletts, a British 
engineer, applied jointly for a British patent for "A New and Im- 
proved Method of and Means for Simultaneously Recording and 
Reproducing Movements and Sounds." British patent 22 was 
granted them on February 11, 1907.* 

FIG. 7. Lauste's "sound and scene" camera (1912-14). 

There is a question of just how much assistance the collaborators 
gave Lauste. Some publications have given them extensive credit, 
while others hold that Lauste had disclosed and conceived the pri- 
mary factors prior to his association with them. Undoubtedly they 
offered some advisory assistance besides their financial aid. 23 ' 24 ' 25 ' 26 

*The patent describes a vibrating galvanometer, oscillograph, light-valve, and 
a projector having a loop between the sound and picture gates. 

April, 1941] 



Lauste continued his experiments with recorders of various types 
until 1910, when he hit upon the idea of using a string galvanometer 
(Figs. 5 and 6). The galvanometer recorder consisted of a mirror 
attached to a silicon bronze wire stretched between the poles of 
two magnets. As current from the microphone passes through the 

FIG. 8. Lauste's "sound and scene" projector 
1912. The lower lamp house was used to illuminate 
the sound-track. Film projector is the old "Pathe 
Professional." This apparatus was first used by 
Lauste in 1911. 

electromagnets, the fluctuating magnetic field caused the string and 
mirror to vibrate accordingly. Movement of the mirror in a beam 
of light traced the sound pattern photographically upon motion 
picture film. Within a year Lauste had made a comparatively suc- 
cessful sound and picture record. The sound record was of the 
variable-width type and occupied half the width of the film; the 
picture utilized the other half (Figs. 7 and 8). The work was con- 

428 W. E. THEISEN [J. S. M. P. E. 

tinued until the War (1915-16), when he largely discontinued his 
researches. These experiments were the foundation for and a 
direct approach to the present sound motion picture. 

Even though the phonograph and gramaphone were considerably 
improved, there were still many factors that hindered its adoption. 
One disadvantage that was particularly noticeable during the period 
before 1910, when there were a great variety of sound devices, was 
that the longest picture that could be recorded was about two hun- 
dred feet, due to limitation of size of record. Most talking pictures 
were only one hundred feet long. 

The maximum playing time of a twelve-inch disk record at that 
time was five minutes, and for the more universally used ten-inch 
records, the time was only three or four minutes. The ten-inch 
record, as used by Cinephone in 1909, accommodated about one 
hundred and fifty feet of film. 

David S. Hulfish, in the Cyclopedia of Motion Picture Work pub- 
lished by the American Technical Society, in Chicago, in 1914, says: 

"The Graphophone has not been perfected for the minor sounds of nature. 
The human voice is about the limit for the sound record. Voices and musical 
instruments are the standard repertoire of the talking machine, other records 
being the exception rather than the rule. This limitation of the Graphophone 
limits the combination sight-and-sound entertainment to dramatic and vaudeville 
incidents, dancing and singing." 

Later in the chapter he says : 

"In producing a talking picture where several actors are involved, the method 
of manufacture is to make the talking record first, and then fit a motion picture 
to it. To do this, the actors are well drilled in their parts, so that they will be 
able to produce the performance twice, once in sound for the sound record and 
once in action for the picture record." 


Unaware of the problems of the talking picture, Theodore W. Case 
began a series of experiments in 1911 that were to be far reaching. 
On January 22nd, Case wrote to his mother: 

"Most of my time is now taken up in experimenting with a selenium cell with 
the idea in mind of photographing sound-waves and using the positives as records 
for a new kind of phonograph, or, rather, it would be called a lithograph, I sup- 


On February 12, 1911, he wrote: 

"Yesterday I at last succeeded in transmitting sound by light. I used the 
principle of the manometric flame." 

Although Case was interested at that time in the problem of photo- 
graphing sound, he did not conduct his experiments with the idea of 
combining motion pictures with sounds. Shortly thereafter his 
experiments in photographing sound were laid aside, and other 
things attracted his interest. In 1916, he began researches that 
were to contribute years later to the development of the talking 
picture photoelectric cell, in the electrical characteristics of many 
materials. By 1917 he had filed an application for a patent on a 
new substance (bismuth-sulfur compound) having variable electri- 
cal resistance in relation to the intensity of light striking it. The 
patent was granted July 8, 1919. 27 ' 28 

The photoelectric cell 29 was the solution of one of the major prob- 
lems of the sound pioneers. With it they could reproduce the 
recorded sound-track in sufficient volume. The first Case Photo- 
electric Cell developed in the Case Research Laboratory at Auburn, 
N. Y. (1913), was known as the thalofide (often spelled tkallafide) 
cell. 30 ' 31 ' 32 There were, of course, other contemporary photoelec- 
tric cells, but they had little use, and, according to available data, 
were used to no great extent during this experimental stage of 
the talking picture. 33 


In the meantime Lee de Forest had been working on a sound 
system. By 1919 he had filed his first patent on a glow-lamp, which 
he later called the "photion" tube, and which he used in his sound 
system. It was the forerunner of the tubes that are now used in 
recording the sound by the "flashing-light" system, which was later 
developed by Case as the aeolight, and introduced as the Fox-Case 
system. Whereas the Case thalofide cell could reproduce the photo- 
graphically recorded sound-track, the de Forest photion tube served 
for recording the original sound record photographically. That 
meant that successful motion pictures and sound could be recorded 
side by side upon the same film, and that they could be reproduced. 

In the next few years de Forest was granted and assigned thirty- 
five patents, 34 and by 1923 he had given several successful demon- 
strations of his phonofilm, among which was a demonstration to the 
press on March 13, 1923, at his New York studio. According to a 



[J. S. M. P. E. 



Operated in conjunction with THE RIAI.TO, Times Square 

KIKE NOTICK-I.ook around NOW and choose the nearest exit to 
yi.ur neat, lu ca.e of flre walk (not run) to THAT EXIT. Do not 
try to beat your neighbor to the street. 

THOMAS J. DREXNAN. Fire Commissioner. 

P R O G R A M 


Evening prices 
prevail at Satur- 
day. Sunday and 
Holiday Matinees 


EVENINGS OrchertraSSwntoj^ 1 ^" 7 J Jg 

Admission 45 cento 

War Tax 5 cento 

Admission 45 cento 

War Tax 5 cento 

Admission 17 cento 

War Tax 3 cento 

Alwiva BO rn Admission 90 CCntS 

Always 99 cents Waf TQX , cent> 

Balcony 50 cento 
MATINEES Orchestra 50 cento 

Balcony 80 cento 


The Rlvoli open at noon dally, performances begin at 12 noon. 2:15*. 
4:13. 7:30* and 9:30 p. m. .Sunday first performance start* at I. 
Asterisks Indicate full ^presentation with orchestra. soloist* and neente 
effect!*. J. Van (left Cooper nnd Frank Stewart Adams at the organ. 


Wi-t-k of April IMh. 1!23 

FRKDKRICK STAIH.nF.RG. Conductor WIL1.IP. STAIIL. Asst. Conductor 

I. OvKirrniK 


Jacques OflVnh.-irh 

FHKDKKICK STAIM ii:m; and WII.I.IK Si AIM. condudtitifr 

Although OrphtMis Is not nuMitioncd in the Homeric poems. the 
story of hi* attempting to hring hark his do id wife, Knrydico. 
from the lower regions was n favorite of the Greeks and has 
U-eii ever since. It w is a subject of the flrst grand opera ever 
written, by Peri in lAoii. and is the theme of Cluck's most famous 
opera, produced in I7*. Offenbach's burlesque of the old myth 
was produced in Paris in IM.I*. and its jolly tune made it a great 
success. The overture is one of I ho finest specimens of light music. 

'2. TIIK Piioxoni.M 

'I'lir I'luiHo/iliii i* the latent ini'eiition of l>r. Lee tie l-'ore*t 
tt ml in n IOIHJ utriile f<ir*'<tr<l in the ilereloinneiit of Hie hilk- 
in>i iiif-hire*. For the jirtl time it lnt* heen imuie />*.*.// 
to r*<-or<l the jiii-ture ami the i"oi<-e or innnir on the X/IMI* 
film, tint* in.iiiriiiif perfect m/iirhroitiznfioii. Tliix ix tjte first 
inililii" nlio\finij. 

FIG. 9. Section of Rivoli Theater (N. Y.) program for 
the week of April, 15, 1923, announcing the de Forest 
phonofilm demonstration. 

program of the Rivoli Theater (Fig. 9), New York, for the week of 
April 15, 1923: 

"The Phonofilm is the latest invention of Dr. Lee de Forest and is a long stride 
forward in the development of the talking pictures. For the first time it has 
been made possible to record the picture and voice or music on the same film, 
thus insuring perfect synchronization. This is the first public showing." 

April, 1941] 



From here the popularity of his sound-on-film pictures extended 
to other theaters and to other cities. 36 ' 36 ' 37 ' 38 

At this stage, however, the quality of the reproduced sound was 
not of sufficient excellence to interest the public greatly. The sound 
was so incomplete in harmonics, so limited in range of intensity, and 
so unmistakably a "canned" product that no illusion was possible. 
It was only after the Western Electric Company had introduced a 
high-quality sound-on-disk system that the motion picture producers 
became really interested in the application of sound to the motion 

FIG. 10 

FIG. 11 

FIG. 10. (Left} "The Kiss," from an Edison picture of 1895. 

FIG. 11. (Right} Sound-track made by system devised by Ruhmer 

(Scientific American, 1901). 

This product of the Western Electric Company was a result of 
many years of research work in the Bell Telephone Laboratories 
which made available the condenser transmitter, the distortionless 
vacuum tube amplifier, and certain adaptations of the telephone 
receiver and transmitter to fit the graver and pick-up needle. The 
design of a precision system of synchronization took only a relatively 
short time of a small group of workers. 


The talking picture had become a possibility; science had tri- 
umphed. It now depended upon a courageous business man with 
foresight to give the pictures to the public. The man to do it was 



[j. S. M. P. E. 



00 0,0 D Q 


Harry Warner. On the night of August 7, 1926, due to his enter- 
prise, Warner Brothers released their first sound picture, Don Juan, 
starring John Barrymore, opening at the New York Warner Brothers 
Theater on Broadway using a sound system developed by the West- 
ern Electric Company and the Bell Telephone Laboratories and com- 
mercially known as the "Vitaphone." In this first program Will 
Hays appeared on the screen in a talking "short," wherein he 
prophesied that the entire course of screen history was to be altered 
by the advent of sound. It was revolutionized. 

Warner Brothers, however, did not record sound-on-film; instead, 
in 1925 they had become interested in a disk system developed by 
the Western Electric Company, who had also a sound-on-film system. 
Warner Brothers, as well as a number of other studios, had been 
offered the choice of either a sound-on-film or a disk system developed 
in the Bell Telephone Laboratories. 

Amplification and disk recording had been perfected in the Bell 
Telephone Laboratories and had been licensed a number of years 
earlier to the Victor Talking Machine Corporation for use in their 
phonographs. Practically the same phonographic system was used 
by Warner Brothers, and by adding loud speakers and amplifiers to 
this system, it was possible to fill the theater with sound of acceptable 
quality. A synchronous motor was added to the phonographic 
system to keep the camera and the sound records in synchronism 
during the recording. 

The second Warner Brothers picture released with sound sequences 
on January 23, 1927, was the Jazz Singer, with Al Jolson, and the 
first completely synchronized feature picture was The Lights of New 
York, released on July 15, 1928. The cast included Helene Costello, 
Cullen Landis, Mary Carr, Gladys Brockwell, and others. It carried 
dialog through its entire length. 

In 1930 Warner Brothers discontinued disks for the most part, in 
favor of sound-on-film, because of the expense of breakage and ship- 
ping the 16-inch records. Vitaphone was the name coined by the 
Warner Brothers for their sound pictures. 


In the meantime, Theodore Case, who had been making thalofide 
cells for Lee de Forest, had extended his interest in the talking picture 
to the point of developing his own system. He is said to have started 
work on a sound camera employing a modulating oxyacetelene flame 



[J. S. M. p. E. 

as early as 1922. In subsequent experiments he found that a vacuum 
tube that he had developed for his infrared-ray signal system could 
be modulated at low voltages for recording sound. This led to the 
development of the aeolight or "flashing lamp" and to the Fox-Case 
sound system in which William Fox acquired a substantial interest 
on September 20, 1926. 

The first commercial exhibition made by Fox-Case of Movietone 
sound, which at this time was the result of combining Western Elec- 
tric Company and Case features, was a short featuring Raquel Meller, 
who sang a number of songs. It was released in conjunction with 
What Price Glory on January 21, 1927. The system was again 
demonstrated in New York on February 25, 1927. On May 29, 

FIG. 13 


FIG. 14 

FIG. 13. Specimen of de Forest phonofilm (1911). 
FIG. 14. Multiple track made by E. H. Amet (1922). 

1927, the first complete Movietone program was shown in the Janet 
Gaynor and Charles Farrell picture, Seventh Heaven. 

The first outdoor sound picture was the Fox Movietone feature, 
In Old Arizona , starring Warner Baxter, first shown at the Los Angeles 
Criterion Theater, December 25, 1928. The first issue of Movietone 
News was released at the Roxy Theater, New York, on April 25, 

1928. Although de Forest had made a few news subjects earlier, 
this was the first talking newsreel. 40 On January 5, 1927, Vitaphone, 
developed by Western Electric, was cross-licensed with Fox-Case 
Movietone. Under this arrangement the Western Electric Com- 
pany manufactured equipment for both systems. 

In the meantime, the Bell Telephone Laboratories had developed 
the light-valve system of recording as contrasted to the flashing 
lamp system of Fox-Case. 



FIG. 15 

FIG. 16 

FIG. 17 

FIG. 18 

FIG. 15. Early photophone recording. 
FIG. 16. Movietone synchronizing mark. 

FIG. 17. Showing a method employed in Germany of producing sound 
effects (1922). Notes at the bottom of each frame served as a guide for 
musicians who sat in the orchestra and sang the music. With earlier 

pictures, the musicians sang behind, or beside, the screen. 
FIG. 18. Sound-track along outer edge of the film (from Germany). 

436 W. E. THEISEN [j. s. M. P. E. 

In the flashing-lamp system the sound record is made by recording 
photographically the variations of intensity of a lamp. These 
variations result from a fluctuating current received through an 
amplification system from a microphone. On the other hand, in the 
light-value method a constant source of light is used, and the sound 
pattern is made by varying the width of a narrow aperture. This 
aperture is formed by a loop of duralumin ribbon stretched between 
magnets so that the two strands of the loop are parallel and about 
0.001 inch apart. When the modulated current from the microphone 
passes through the ribbons, the distance between the two strands 
varies. These variations when photographed result in a sound record 
of the variable density type. 

This development of the Bell Telephone Laboratories was the re- 
sult of the efforts of a large number of workers, particularly Edward 
C. Wente and Donald MacKenzie. Wente developed the condenser 
transmitter for the telephone. The condenser microphone was used 
also in the de Forest system, and has been applied extensively in the 
recording of sound pictures in recent years. Wente was awarded the 
Progress Medal of the Society in 1935 in recognition of his researches. 
MacKenzie was largely responsible for coordinating the various ele- 
ments and for working out the necessary conditions for processing 
variable-density sound records. The contributions of other telephone 
engineers should not be overlooked. In their many years of practical 
experience and research into the problems of electrical communica- 
tion circuits they have contributed much to the talking picture. 
Included in this group is H. B. Wier, who was among the first to under- 
stand the advantages of re-recording and the use of equalization 
networks. On the commercial side, and in the organization of de- 
velopments in the Bell Telephone Laboratories, the steady support 
given by Edward B. Craft, Chief Engineer, was extremely valuable, 
and perhaps served as the most stimulating force in maintaining the 
interest of this research organization.* 

By May 15, 1928, Metro-Goldwyn-Mayer, Paramount-Famous 
Players-Lasky Corporation, and United Artists Studio had identified 
themselves with Western Electric, using the light- valve system. 

This survey in sound pioneering from this point concerns itself largely with 
the commercial application of the various sound systems rather than detailing 
the improvements and contributions of the engineers in the laboratories. Their 
work is of inestimable value to the present high quality of sound engineering 
and the record of their work remains to be written. 


Very shortly thereafter, First National* Christie, and Universal 
Pictures were also licensed by Western Electric. 

Metro-Goldwyn-Mayer produced the first short in color and sound, 
namely, Gus Edward's Color-Tone Revue, first shown at the Carthay 
Circle Theater, Los Angeles. The first feature having color (two- 
color Technicolor) as well as sound was the Metro-Goldwyn-Mayer 
Broadway Melody, released on February 17, 1929. It as well as 
most other pictures of the time, was made in both silent and talkie 

FIG. 19. 42-mm. track with the 
sound outside the perforations (from 

The necessity of making the two versions may be seen from the 
fact that by September 27, 1928, only six hundred theaters were 
wired for sound by Western Electric. At that time there were 
plans to wire another four hundred by January 1st, and an additional 
2000 during 1929. 


Installations of RCA Photophone equipment were begun on 
October 1, 1928. 41 Photophone, a system developed by the Radio 
Corporation of America, Westinghouse Electric & Manufacturing Co., 
and General Electric Company, entered the sound-picture situa- 
tion January 7, 1928, when a substantial interest was acquired in 
Film Booking Offices (known in Hollywood then as F. B. O. Studios). 
In March of that year negotiations were begun for the acquisition 



[J. S. M. P. E. 

of the Victor Talking Machine Corporation, and were consummated 
on January 6, 1929. The system with improvements was intro- 
duced as the Photophone February 1, 1928, and was demonstrated 
before the Society at Hollywood in April, 1928, by H. B. Marvin. 42 
The first Photophone feature was the F. B. O. picture, The Perfect 
Crime, released on June 17, 1928. 

Charles A. Hoxie of the General Electric Company, who was 
responsible for much of the pioneering research in the Photophone 
system, started working with the sound and photographic problem in 

FIG. 21 

FIG. 20 

FIG. 20. Specimen of cinephone recording of 1928, used with Mickey 

Mouse cartoons. 
FIG. 21. Recent Mickey Mouse recording. 

1920. The device he constructed at the time was known as the 
pallophotophone. His first camera for recording sound was made in 
January, 1921, with which he recorded speeches of President Coolidge, 
the Secretary of War, and others. The speeches were broadcast 
over radio station WGY in 1922. This led directly to the commer- 
cial models of the photophone equipment. A demonstration includ- 
ing the Volga Boatman was given before motion picture producers at 
Schenectady in 1926, and, as a sound and picture device, the kinegra- 
phone, as it was termed, was demonstrated at the State Theater in 
Schenectady, N. Y., in September, 1927. An exhibition including 
reels of Flesh and the Devil was given at the Rivoli Theater in New 
York early in 1927. 


In contrast to the variable-density type of record, the RCA 
Photophone method of recording sound in the earlier systems was 
somewhat similar in basic principle to the Lauste method in 1910. 
A constant source of light was used, which played upon a mirror 
attached to wire ribbons stretched between electromagnets. When 
the electromagnets received the electrical impulses from a microphone 
the mirror was deflected. The movement of the light-beam reflected 
from the mirror was photographed, producing what is known as a 
variable- width sound-track. 

The system has been improved recently by substituting for the 
wire ribbons an armature, which turns upon an axis between the 
poles of an electromagnet, the mirror being attached to the armature. 
The voice currents from the microphone energize the electromagnet 
which in turn deflects the armature and mirror. Some of the earli- 
est RCA Photophone recordings were the RKO Dixiana, the Pathe- 
de Mille King of Kings, and The Godless Girl. 

Research work devoted to the problem of extending the frequency 
range and reducing distortion in both recording and reproducing 
systems conducted by the research engineers of the RCA Victor 
Company resulted in the development of a commercial system having 
an improved signal-to-noise ratio, increased frequency range, and 
reduced harmonic distortions. The RCA Photophone high-fidelity 
system was announced in 1932. Corresponding improvements 
were also made in the Western Electric System. 

The improvements in the Photophone system were a result of the 
efforts of a large number of research workers located at Camden, 
N. J., after the Photophone interests were absorbed from the General 
Electric Company by the RCA Corporation. Most noteworthy 
among these workers were M. C. Batsel, G. L. Dimmick, E. W. 
Kellogg, and E. W. Engstrom. Kellogg was awarded the Progress 
Medal of the Society in 1937 in recognition of his researches. 

Patents on noise-reduction were also granted to Siemens and 
Halske. 43 In referring to the experimental work of the German sound 
engineers, it should be noted that Ernst Vorbeck of Berlin proposed 
the idea of using electric amplifiers for sound registration on March 26, 
1913 (D. R. P. 285,492). On May 4, 1919, Hans Vogt, Jo Engl, 
and Joseph Massolle filed the first of a series of patents covering sound 
recording (D. R. P. 417,967). Engl had first become interested in 
what he called "talking Pictures" in 1912. These three engineers 
perfected the Tri-Ergon process, which was adopted by "Tobis" and 



[J. S. M. P. E, 

the first public demonstration was held at the Apollo Theater in 
Paris in May, 1929, under the direction of Dr. Hans Henkel. 44 - 45 

Besides the two major systems, developed in the United States, 
the Western Electric Movietone and the RCA Victor Photophone, 
there were a large number of systems of lesser importance: notably 
the Cinephone system, developed by Pat Powers, R. I. Halpenny, 
and William Garity, which was distinguished by being the system 
used for recording the first sound cartoon, a Mickey Mouse picture 

FIG. 22 

FIG. 24 

FIG. 22. Split 35-mm. recording used by Metro-Gold wyn-Mayer and Uni- 
versal as an economy measure. First used by M-G-M in 1931. 
FIG. 23. RCA 16-mm. sound-film, announced in spring of 1932. 
FIG. 24. High-fidelity RCA recording, announced in May, 1932. 

entitled Steamboat Willie. This cartoon, which incidentally, was 
the first of the Mickey Mouse series, was released at the Colony 
Theater, New York, in September, 1928. The Cinephone was basi- 
cally the de Forest phonofilm system. 

Among other systems that had some use were the Hanaphone, 
Vocafilm, Biophone, Kolstatone, Milotone, Reeltone, and Phono- 
scope. There were more than eighty film and disk systems put out 
by different companies; most of them being based upon the Western 
Electric or RCA Photophone principles. For recording there were 
ten disk and five sound-film systems available. 46 ' 47 - 48 ' 49 


This number of systems brought forth the question of standardi- 
zation and interchangeability. On July 7, 1928, this question first 
arose when Western Electric Company gave permission to RCA 
Photophone to show King of Kings on Western Electric equipment 
at the Rivoli Theater, New York. Photophone had been using a 
wider sound-track (100 mils) than was usual with the variable- 
density type of record, until July 17, 1928, at which time they adopted 
the 80-mil width as used with the latter. This made it possible 
to reproduce photophone recordings in the comparatively large 
number of theaters equipped with Western Electric apparatus. 
Theater installations of RCA Photophone equipment began in Octo- 
ber, 1928. 

The question of interchangeability was largely settled on December 
30, 1928, when J. E. Otterson, then President of Electrical Research 
Products, Inc., a subsidiary of the Western Electric Company, 
announced that an agreement had been reached among licensees of 
the various systems. 

Fox Studios was one of the first companies to discontinue making 
dual versions. Beginning March 1, 1929, silent versions were dis- 
continued and only talking pictures were made thereafter. This 
decision was somewhat daring, since a large number of the leaders 
of the industry still felt that sound-films were only a passing fad. 
In 1929, 348 talking and 400 silent pictures were made. Metro- 
Goldwyn-Mayer in November of that year began making foreign 
versions. Europe, in the meantime, was following a more conserva- 
tive plan in making talking pictures. 

The sound-film had become established, and with it came a new 
industry. The silent motion picture with its dramatic devices were 
in the discard, and in its place was an entirely new screen medium. 
In the scramble of 1928-29, a new dramatic form was evolved. The 
entire motion picture had changed. 50 - 51 ' 62 ' 53 

Reproductions of specimens of film showing the various forms of 
sound-track evolved in the development of the talking picture are 
shown in Figs. 10-24, enlarged somewhat in order to show the sound- 
tracks clearly. The original specimens are in the SMPE motion 
picture exhibit at the Los Angeles Museum. 

442 W. E. THEISEN [J. S. M. P. E. 


1 CASSELL AND Co: "Encyclopedia of Photography," 14th ed., 1914, p. 166. 

2 HOPWOOD, H. V.: "Living Pictures," Optician & Photographic Trades Re- 
view (London), 1899, p. 62. 

8 DEMENY, G: "Les Photographies Parlantes" (phonoscope), La Nature, 
1892 Part I, p. 311. 

4 DICKSON, W. K. L. : "A Brief History of the Kinetograph, Kinetoscope, and 
Kinetophonograph," J. Soc. Mot. Pict. Eng., XXI (Dec., 1933), No. 6, p. 435. 

6 SEiBT, G. (Ruhmer's associate): D. R. P. No. 418,051; March 11, 1919. 

6 Phys. Zeit. (May 6, 1901), No. 34, p. 498. 

7 Booklet (received by author Dec., 1933) : Gaumont, L. : "Etablissments 
Gaumont, 1895-1929," Gauthier-Villars (Paris), 1936. 

8 Bull. Soc. Franc,, de Phot. (2nd Series), XVIII (Nov. 7, 1902), No. 22, p. 500. 

9 Ibid. (3rd Series), II (Feb. 17, 1911), No. 3, pp. 96 and 111. 

10 "L'Industrie du Film Parlant," Conservatoire des Arts et Metiers (Feb. 17, 
1929). (Contains a resume covering the Gaumont and the Gaumont-Peterson- 
Poulson sound systems.) 

11 La Cinemat. Franc. (May 27, 1933), No. 760, p. 25. 

12 Brit. Pat. No. 22,563; No. 22,564; No. 22,565. 

18 FRANKLIN, H. B.: "Sound Motion Pictures," Doubleday, Doran & Co. 
(New York), 1929, p. 5. 

14 Brit. Pat. No. 10,543 (May 4, 1909). 

Brit. Pat. No. 10,417 (April 28, 1910); No. 10,779 (May 2, 1910). 

16 HOPWOOD AND FOSTER: "Living Pictures," Hatton Press (London), 1915, 
p. 274. 

17 HULFISH, D. S.: "Cyclopedia of Motion Picture Work," Amer. Tech. Soc., 
I (1914), p. 241. (Containing technical details of magnetic and escapement 
synchronizing devices.) 

18 U. S. Pat. : "Method and Means for Localizing Sound Reproduction," No. 
1,124,580 (Jan. 12, 1915). 

U. S. Pat. : "Combined Phonograph and Motion Picture Apparatus for Repro- 
ducing Indexed Synchronous Records," No. 1,162,433 (Nov. 20, 1915). 

19 U. S. Pat.: "Balanced Electrical Transmitter," No. 1,176,725 (May 21, 

U. S. Pat.: "Means for Reproducing Vibrations," No. 1,221,408 (April 3, 

20 U. S. Pat.: "Apparatus for Preparing Combined Cinematographic and 
Phonographic Records," No. 1,204,091 (Nov. 7, 1916). 

21 U. S. Pat.: "Automatically Synchronizing Entertainment Devices," No. 
1,234,127 (July 4, 1917). 

BRISTOL, W. H.: "An Electrical Synchronizing and Resynchronizing System 
for Sound Motion Picture Apparatus," Trans. Soc. Mot. Eng., XII (1928), No. 
35, p. 778. 

22 Brit. Pat. No. 18,057 (Feb. 11, 1907). 

23 CRAWFORD, M. : "Pioneer Experiments of Eugene A. Lauste in Recording 
Sound," /. Soc. Mot. Pict. Eng., XVII (Oct., 1931), No. 4, p. 632. 

24 "Talking Pictures," Photo-Era, XXIII (Dec., 1909). 


26 Photo-Era, XXIII (Nov., 1909), p. 228. 

26 "Talking Films," Sydney (Australia) Morning Herald (Aug. 26, 1929). 

27 U. S. Pat.: No. 1,309,181 (another patent relating to a photosensitive sub- 
stance was granted Sept. 16, 1919). 

28 U. S. Pat.: "New Compound Showing a Variable Resistance under the In- 
fluence of Light," No. 1,316,220. 

29 DYSART, P. M. : Description of the use of the glow-lamp in an article in 
"School Science and Mathematics," Smith & Turton (Mt. Morris, 111.), XIV 
(Jan., 1914), No. 1, p. 36. 

30 CASE, T. W.: "Thalofide Cell, a New Photoelectric Substance," Phys. 
Rev., 15 (April, 1920), p. 298. 

81 "New Advances Made in Talking Movies," Yale Sci. Mag. (May, 1927). 

32 "Thalofide Cell," Bull. No. 4, Case Research Laboratories, 1921. 

33 TYKOCINSKI-TYKOCINER, J.: "Photographic Recording and Photoelectric 
Reproduction of Sound," Trans. Soc. Mot. Pict. Eng., VII (1923), No. 16, p. 90. 

34 Notable de Forest patents: U. S. Pat. No. 1,482,246 (filed Sept. 18, 1919; 
issued March 11, 1924). 

U. S. Pat. No. 1,482,119 (filed Sept. 18, 1919; issued Jan. 29, 1924). 
36 MILLS, J.: "Signals and Speech in Electrical Communication," Har court, 
Brace & Co. (New York), 1934, p. 166. 

36 DE FOREST, L.: "The Phonofilm" Trans. Soc. Mot. Pict. Eng., VII (1923), 
No. 16, p. 61. 

37 DE FOREST, L.: "Phonofilm Progress," Trans. Soc. Mot. Pict. Eng., VIII 
(1924), No. 20, p. 17. 

38 DE FOREST, L.: "Recent Developments in the Phonofilm," Trans. Soc. 
Mot. Pict. Eng., X (1926), No. 27, p. 64. 

39 Film Daily Year Book, llth ed. (1929), p. 253. 

40 Ibid., p. 198. 

41 Ibid. p. 561. 

42 MARVIN, H. B.: A System of Motion Pictures with Sound," Trans. Soc. 
Mot. Pict. Eng., XII (1928), No. 33, p. 86. 

43 Brit. Pat. No. 288,225, granted to Siemens & Halske (Convention date, 
April 9, 1927); published in Illustrated Official Journal Patents, May 31, 1928, 
p. 2114; granted Jan. 10, 1929. 

44 Ruhmer was later associated with Lauste. Other patents were: D. R. P. 
No. 387,058 (May 23, 1920). 

D. R. P. No. 368,383 (April 15, 1921); U. S. Pat. No. 1,825,598. 
American Court of Customs and Patent Appeals rendered decision No. 2749, 
May 25, 1931, regarding priority on the similar de Forest patents. 
46 Massole in 1930 invented a portable Tobis sound apparatus. 

46 Film Daily Year Book, 12th ed. (1930), p. 873. 

47 ENGL, J. B.: "A New Process for Developing and Printing Photographic 
Sound Records," Trans. Soc. Mot. Pict. Eng., XI (1927), No. 30, p. 257. 

48 TEUKE, K.: "Story of Sound-Film," Kinotechnik, 12 (Feb. 5, 1930), p. 65. 

49 "Ries Sound-Film Patents," Mot. Pict. News, 42 (July 5, 1930), p. 21. 

60 CRAWFORD, M.: "Evolution of Sound Pictures," Bull Internal. Phot.. 
(March, 1930). 

444 W. E. THEISEN 

61 HULFISH, D. S. : "When the Movies First Tried to Talk," Examiners Herald 
World (Aug. 30, 1930), p. 33. 

52 GREEN, F. : "The Film Finds Its Tongue," G. P. Putnam's Sons (New York) , 

63 PITKIN AND MARSTON, L. : "The Art of Sound Pictures," D. Appleton & Co. 
(New York), 1930. 


ENGLE, J.: "Der Tonende Film," F. Vieweg & Son (Braunschweig, Germany), 

VON MIHALY, D.: "Der Sprechende Film," M. Krayn (Berlin), 1927. 

UMBEHR, H.: "Der Tonfilm" (ed. by H. Wollenberg), Lichtbildbuhne (Berlin), 
1930; Vol. 4, Bucher der Praxis. 

"Recording Sound for Motion Pictures" (ed. by L. Cowan), McGraw-Hill Book 
Co. (New York), 1931. 

FISCHER, F., AND LICHTE, H. : "Tonfilm-Aufnahme und Wiedergabe nach dem 
Klangfilm-Verfahren," S. Hirzel (Leipzig), 1931. 

BROWN, B.: "Talking Pictures," I. Pitman & Son (London), 1931. 

MILLER, D. C. : "Anecdotal History of the Science of Sound," MacMillan & 
Co. (New York), 1934. 



Engineering Vice-President Executive Vice-President 

Editorial Vice-P resident 


Financial Vice-President 


Convention Vice-President 













Chairman, Atlantic 

Coast Section 


Chairman, Mid- West 


Chairman, Pacific 

Coast Section 


(Atlantic Coast) 

*R. O. STROCK, Chairman 

*P. J. LARSEN, Past-Chairman **W. H. OFFENHAUSER, JR., Manager 

*J. A. MAURER, Sec.-Treas. *H. B. CUTHBERTSON, Manager 

**P. C. GOLDMARK, Manager *J. A. NORLING, Manager 

**H. E. WHITE, Manager *F. E. CAHILL, Manager 


*J. A. DUBRAY, Chairman 

S. A. LUKES, Past-Chairman *C. H. STONE, Manager 

"I. JACOBSEN, Sec.-Treas. *G. W. COLBURN, Manager 

(Pacific Coast) 

*J. G. FRAYNE, Chairman 

*L. L. RYDER, Past-Chairman **S. P. SOLOW, Manager 

*C. W. HANDLEY, Sec.-Treas. *H. W. MOYSE, Manager 

**F. J. DURST, Manager *R. W. REMERSHEID, Manager 

**B. KREUZER, Manager *P. MOLE, Manager 

* Term expires December 31, 1941. 
** Term expires December 31, 1942. 


(Correct to March 20th; additional appointments or changes may be made at any 
time during the year, as necessity or expediency may require.} 



(East Coast) 

G. FRIEDL, JR., Chairman 



(West Coast) 

L. L. RYDER, Chairman 




A. C. DOWNES, Chairman 







R. M. EVANS, Chairman 


W. C. KUNZMANN, Chairman 


O. F. NEU 




I. D. WRATTEN, Chairman 



J. S. MACLEOD, Chairman 





J. FRANK, JR., Chairman 


E. A. WILLIFORD, Chairman 




R. E. FARNHAM, Chairman 




H. E. WHITE, Chairman 





J. FRANK, JR., Chairman 









J. A. MAURER, Chairman 










S. HARRIS, Chairman 
G. A. CHAMBERS, West Coast Chairman 

[J. S. M. P. E. 





J. G. BRADLEY, Chairman 










F. C. GILBERT ; Chairman 




K. F. MORGAN, Chairman 




J. HABER, Chairman 




H. G. TASKER, Chairman 







April, 1941] 







D. B. JOY, Chairman 














R. LINDERMAN, Chairman 











P. C. GOLDMARK, Chairman 




A. N. GOLDSMITH, Chairman 

(Projection Practice Sub- Committee) 

H. RUBIN, Sub-Chairman 









(Theater Design Sub-Committee) 

B. SCHLANGER, Sub-Chairman 


J. FRANK, JR. E. R. MORIN R. F. Ross 

(Screen Brightness Sub- Committee) 

F. E. CARLSON, Sub-Chairman 





American Documentation Institute J. E. ABBOTT 

Sectional Committee on Motion Pictures, ASA E. K. CARVER 


Sectional Committee on Photography, ASA J. I. CRABTREE 

Inter-Society Color Council R. M. EVANS 



Sectional Committee on Standardization of Letter L. A. JONES 

Symbols and Abbreviations for Science and Engi- 
neering, ASA 














* Alternate 



The editors present for convenient reference a list of articles dealing with subjects 
cognate to motion picture engineering published in a number of selected journals. 
Photostatic copies may be obtained from the Library of Congress, Washington, D. C., 
or from the New York Public Library, New York, N. Y. Micro copies of articles 
in magazines that are available may be obtained from the Bibliofilm Service, Depart- 
ment of Agriculture, Washington, D. C. 

American Cinematographer 

22 (March, 1941), No. 3 
Putting Naturalness into Modern Interior Lightings 

(pp. 104-105, 136) 

Growing Pains (pp. 106-107, 139-142) 
Aces of the Camera. Ill: Ray Rennahan, A.S.C. (pp. 

108, 134, 136) 

Movies for National Defense (pp. 109, 130, 132, 134) 
Innovations in New Williams Laboratory (pp. 110, 134) 
Surgical Cinematography (pp. 120-121) 
Projecting Sound and Silent Film (pp. 122-123, 149-150) 

Acoustical Society of America, Journal 

12 (January, 1941), No. 3 

A Theory of Tracing Distortion in Sound Reproduction 
from Phonograph Records (pp. 348-365) 

The Analysis of Pulses by Means of the Harmonic Ana- 
lyzer (pp. 383-386) 

The Influence of Certain Atmospheric Conditions upon 
Sound Transmission at Short Range (pp. 427-435) 

Educational Screen 

19 (February, 1941), No. 2 
Motion Pictures Not for Theaters (pp. 61-64), Pt. 24 


14 (February, 1941), No. 2 
Sound in Motion Pictures (pp. 37-39, 88), Pt. II 

International Projectionist 

16 (January, 1941), No. 1 

Modern Microphone Types, Structure and Operating 
Technic (pp. 7-8, 13) 












Resonant Circuits (pp. 14-16) 

A New Treatment for the Prevention of Film Abrasion 

and Oil Mottle (pp. 17, 34) 
RCA's Theater Tele Show on 15 X 20-Ft. Screen an 

Historic Event in the Electronics Art (pp. 18, 21) 
Fundamentals of Sound Reproduction (pp. 27-29) 

Photographische Industrie 

39 (January 2, 1941), No. 1 

Lautstarkeumfang der Tonfilmwiedergabe (Range 
Volume Level in Sound Reproduction) (pp. 14-15) 








Program and Facilities 

E. HUSE, President 

E. A. WILLIFORD, Past-President 

H. GRIFFIN, Executive Vice-P resident 

W. C. KUNZMANN, Convention Vice-P resident 

A. C. DOWNES, Editorial Vice-President 

G. A. BLAIR, Chairman, Local Arrangements 

S. HARRIS, Chairman, Papers Committee 

J. HABER, Chairman, Publicity Committee 

J. FRANK, JR., Chairman, Membership Committee 

H. F. HEIDEGGER, Chairman, Convention Projection 

Reception and Local Arrangements 





G. A. BLAIR, Chairman 
C. E. K. MEES 

Registration and Information 

W. C. KUNZMANN, Chairman 

Hotel and Transportation 

F, E. ALTMAN, Chairman 

Publicity Committee 

J. HABER, Chairman 








[J. S. M. p. E. 




Banquet and Dance 

I. L. NIXON, Chairman 

Ladies' Reception Committee 

MRS. C. M. TUTTLE, Hostess 


H. F. HEIDEGGER, Chairman 







Officers and Members Rochester Projectionists Local No. 253 


The headquarters of the Convention will be the Sagamore Hotel, where excellent 
accommodations and moderate rates are assured. A reception parlor will be pro- 
vided as headquarters for the Ladies' Committee. 

Hotel reservation cards mailed to the members of the Society several weeks 
ago should be filled out and mailed immediately to the Sagamore Hotel so that 
suitable accommodations may be reserved, subject to cancellation if unable 
to attend the Convention. 

The following European-plan day rates are extended by the Sagamore Hotel to 
Society members and guests attending the Convention (all rooms are outside 
rooms with bath) : 

Room for one person 

Room for two persons, double bed 

Room for two persons, twin beds 

Suite accommodations, one to four persons 

$3. 00 to $5. 00 
4. 50 to 6.00 
6. 00 to 7.00 

12.00 and up 

The following hotel garage rates will be available to SMPE delegates and guests 
who motor to the Convention : 24-hr, inside parking, 75 ; outside parking (daily), 

The colorful Sagamore Room on the main floor of the Hotel offers special break- 
fast, luncheon, and dinner menus at moderate prices. 

Golfing privileges at several Rochester country clubs may be arranged for 
either by the hotel management or at the SMPE registration headquarters. 

April, 1941 ] SPRING CONVENTION 457 


Convention registration and information headquarters will be located on the 
Sagamore Hotel roof, adjacent to the Glass House, where technical sessions and 
symposiums will be held. 

Members and guests attending the Convention will be expected to register, and 
so help to defray the Convention expenses. Convention badges and identifica- 
tion cards will be provided for admittance to all regular and special sessions during 
the Convention. The identification card will also be honored through the 
courtesy of Loew's Theaters, Inc., at Loew's Rochester Theater and, through the 
courtesy of Monroe Amusements, Inc., at the Palace, Regent, and Century Theaters. 


All the technical sessions of the Convention will be held hi the Glass House 
on the roof of the Sagamore Hotel with the exception of Wednesday morning and 
evening, as described below. Members should note that the banquet, which 
at past conventions has always been held on Wednesday evening, this tune has 
been scheduled for Tuesday evening to permit holding a special meeting on 
Wednesday evening at the Eastman Theater. 

Wednesday, May 7th, will be devoted to a joint meeting of the Acoustical So- 
ciety of America and the SMPE, consisting of a symposium of papers by engineers 
of the Bell Telephone Laboratories in the morning and afternoon. In the evening 
a demonstration of stereophonic sound will be given by the Bell Telephone Labora- 
tories at the Eastman Theater. 


The usual Informal Get-Together Luncheon for members, their families, and 
guests will be held in the Starlight Room on the hotel roof on Monday, May 5th, 
at 12:30 P. M. Luncheon and banquet tickets should be procured when register- 

The 48th Semi-Annual Banquet and Dance will be held in the Starlight Room 
on the hotel roof on Tuesday evening, May 6th, at 7:30 P. M.: music and enter- 
tainment. Banquet tickets should be procured and tables reserved at registra- 
tion headquarters by noon of Tuesday, May 6th. 


Mrs. C. M. Tuttle, Convention Hostess, and members of her Committee are 
arranging a very attractive program of entertainment for the ladies attending the 
Convention. A reception parlor will be provided for the use of the Committee 
during the Convention. 



Monday, May 5th 
9:00 a. m. Sagamore Hotel Roof 

9:30 a. m.-12:00 Glass House, Hotel Roof 

Technical session 
12:30 p. m. Starlight Room, Hotel Roof 

Get-Together Luncheon for members, their families, and 
guests. Brief addresses by several prominent speakers 
2:00 p. m. Glass House, Hotel Roof 

Technical session 

8:00 p. m. Glass House, Hotel Roof 

Technical session 

Tuesday, May 6th 
9:00 a. m. Sagamore Hotel Roof 

9:30 a. m. Glass House, Hotel Roof 

Society Business 

Technical session 
12:30 p. m. Luncheon period 

Open afternoon 
7:30 p. m. Starlight Room, Hotel Roof 

Semi- Annual Banquet and Dance of the SMPE: addresses 

and entertainment: music, dancing, and entertainment 

Wednesday, May 7th 

Joint Meeting of the Acoustical Society of America and the Society of 

Motion Picture Engineers 
10:00 a. m. Eastman Theater 

Stereophonic sound papers session 
12:30 p. m. Luncheon period 

2:00 p. m. Glass House, Hotel Roof 

Stereophonic sound papers session 
8:00 p. m. Eastman Theater 

Stereophonic sound demonstration for the SMPE Conven- 
tion and invited groups. Admission only by SMPE 
identification card, or special invitation card 

Thursday, May 8th 
10:00 a. m. Glass House, Hotel Roof 

Technical session 

12:30 p. m. Luncheon period 

2:00 p. m. Glass House, Hotel Roof 

Technical session 



Convention Vice- President 



(March, 1941, page 264} 

In the "Report of the SMPE Standards Committee," published in the March, 
1941, issue of the JOURNAL, p. 264, the chart showing the specifications for raw- 
stock cores for 35-mm film was headed "American Standards." 

This was in error, as the chart should have been headed "SMPE Recommended 

The specifications have not yet been submitted to the Sectional Committee on 
Motion Pictures (Z-22) of the American Standards Association, and hence should 
not be regarded as "American Standards." 



MAY 5th-8th, INC. 

Plans for the approaching Rochester Convention are proceeding rapidly and 
the tentative program of presentations will be mailed to the members of the So- 
ciety in the near future. 

Preliminary bulletins and hotel reservation cards have recently been mailed to 
the membership, and all members planning to attend the Convention should be 
sure to return their reservations cards without delay. The Acoustical Society 
of America will meet in Rochester May 5th and 6th, concurrently with the first 
two days of the SMPE Convention, and promptness in making reservations may 
be necessary in order to secure satisfactory accommodations. 

Further details concerning the Convention will be found on page 455 of this 
issue of the JOURNAL. 


At a meeting held at the Hotel Pennsylvania, New York, N. Y., on the evening 
of March 19th, Dr. Harry F. Olsen of the RCA Manufacturing Company, Cam- 
den, N. J., presented a comprehensive talk on the subject of "Microphones for 
Motion Pictures Their Uses, Characteristics and Pick-Up Technics." 

The meeting was unusually well attended and considerable discussion followed 
the presentation. Preceding the meeting, a group of members met for supper 
in the Coffee Shop of the Hotel. 


On February 25th, the regular monthly meeting of the Section was held in the 
meeting rooms of the Western Society of Engineers, Chicago, 111. 

The subject of the meeting, The Alchemist in Hollywood, was a 16-mm sound- 
film showing, in an educational and technical way, the nature of motion picture 
activities in Hollywood. The main subject was prefaced by a short talk by the 
Chairman of the Section. 




The February meeting of the Section was held at the Samuel Goldwyn Studios 
in Hollywood on February 25th. Mr. Roy S. Leonard presented a 16-mm Koda- 
chrome picture produced for. the Lighting Department of Seattle, Wash., entitled 
The Million Horsepower Skagit. The picture presented the story of the scenic 
beauty and the power development of the 1200 square-mile mountain area of the 
Skagit River watershed. 

In addition, Mr. Carl Dunning presented a paper on Making 16- Mm Koda- 
chrome Duplicates, as done in the Dunningcolor Laboratories. The paper was ac- 
companied by a demonstration reel. 

Also, a paper prepared by Mr. L. R. Martin of the Eastman Kodak Company, 
on the subject of The Motion Picture as a Tool in Science and Engineering was pre- 
sented, accompanied by demonstration films. 



Article 1 


The name of this association shall be SOCIETY OF MOTION PICTURE 

Article II 


Its objects shall be: Advancement in the theory and practice of motion pic- 
ture engineering and the allied arts and sciences, the standardization of the equip- 
ment, mechanisms, and practices employed therein, the maintenance of a high 
professional standing among its members, and the dissemination of scientific 
knowledge by publication. 

Article III 


Any person of good character may be a member in any class for which he is 

Article IV 


The officers of the Society shall be a President, a Past-President, an Executive 
Vice-President, an Engineering Vice-President, an Editorial Vice-President, a 
Financial Vice-President, a Convention Vice-President, a Secretary, and a 

The term of office of the President, the Past-President, the Executive Vice- 
President, the Engineering Vice-President, the Editorial Vice-President, the 
Financial Vice-President, and the Convention Vice-President shall be two years, 
and the Secretary and the Treasurer one year. Of the Engineering, Editorial, 
Financial, and Convention Vice-Presidents, two shall be elected alternately each 
year, or until their successors are chosen. The President shall not be immediately 
eligible to succeed himself in office. 

Article V 

Board of Governors 

The Board of Governors shall consist of the President, the Past-President, the 
five Vice-Presidents, the Secretary, the Treasurer, the Section Chairmen, and 

* Corrected to March 15. 1941. 



five elected Governors. Two, and three, of the Governors shall be elected al- 
ternately each year to serve for two years. 

Article VI 


There shall be an annual meeting, and such other meetings as stated in the 

Article VII 


This Constitution may be amended as follows: Amendments shall be approved 
by the Board of Governors, and shall be submitted for discussion at any regular 
members' meeting. The proposed amendment and complete discussion then 
shall be submitted to the entire Active, Fellow, and Honorary membership, 
together with letter ballot as soon as possible after the meeting. Two-thirds of 
the vote cast within sixty days after mailing shall be required to carry the amend- 

By-Law I 


Sec. 1. The membership of the Society shall consist of Honorary members, 
Fellows, Active members, Associate members, and Sustaining members. 

An Honorary member is one who has performed eminent services in the ad- 
vancement of motion picture engineering or in the allied arts. An Honorary 
member shall be entitled to vote and to hold any office in the Society. 

A Fellow is one who shall not be less than thirty years of age and who shall 
comply with the requirements of either (a) or (6) for Active members and, in 
addition, shall by his proficiency and contributions have attained to an out- 
standing rank among engineers or executives of the motion picture industry. 
A Fellow shall be entitled to vote and to hold any office in the Society. 

An Active member is one who shall be not less than 25 years of age, and shall 

(a) A motion picture engineer by profession. He shall have been engaged in 
the practice of his profession for a period of at least three years, and shall have 
taken responsibility for the design, installation, or operation of systems or ap- 
paratus pertaining to the motion picture industry. 

(&) A person regularly employed in motion picture or closely allied work, 
who by his inventions or proficiency in motion picture science or as an executive 
of a motion picture enterprise of large scope, has attained to a recognized stand- 
ing in the motion picture industry. In case of such an executive, the applicant 
must be qualified to take full charge of the broader features of motion picture 
engineering involved in the work under his direction. 

(c) An Active member is privileged to vote and to hold any office in the So- 

An Associate member is one who shall be not less than 18 years of age, and shall 
be a person who is interested in or connected with the study of motion picture 


technical problems or the application of them. An Associate member is not privi- 
leged to vote, to hold office or to act as chairman of any committee, although he 
may serve upon any committee to which he may be appointed; and, when so 
appointed, shall be entitled to the full voting privileges of a committee member. 

A Sustaining member is an individual, a firm, or corporation contributing 
substantially to the financial support of the Society. 

Sec. 2. All applications for membership or transfer, except for honorary or 
fellow membership, shall be made on blank forms provided for the purpose, and 
shall give a complete record of the applicant's education and experience. Honor- 
ary and Fellow membership may not be applied for. 

Sec. 3. (a) An Honorary membership may be granted upon recommendation 
of the Board of Governors when confirmed by a four-fifths majority vote of the 
Honorary members, Fellows, and Active members present at any regular meeting 
of the Society. An Honorary member shall be exempt from all dues. 

(&) Fellow membership may be granted upon recommendation of at least 
three-fourths of the Board of Governors. 

(c) Applicants for Active Membership shall give as reference at least three 
members of Active or of higher grade in good standing. Applicants shall be elected 
to membership by the unanimous approval of the entire membership of the ap- 
propriate Admissions Committee. In the event of a single dissenting vote or 
failure of any member of the Admissions Committee to vote, the application shall 
be referred to the Board of Governors, in which case approval of at least three- 
fourths of the Board of Governors shall be required. 

(d) Applicants for Associate membership shall give as reference at least one 
member of higher grade in good standing. Applicants shall be elected to member- 
ship by approval of a majority of the appropriate Admissions Committee. 

By-Law II 


Sec. 1. An officer or governor shall be an Honorary, a Fellow, or Active mem- 

Sec. 2. Vacancies in the Board of Governors shall be filled by the Board of 
Governors until the annual meeting of the Society. 

By-Law III 

Board of Governors 

Sec. 1. The Board of Governors shall transact the business of the Society 
between members' meetings, and shall meet at the call of the president. 

Sec. 2. A majority of the Board of Governors shall constitute a quorum at 
regular meetings. 

Sec. 3. When voting by letter ballot, a majority affirmative vote of the total 
membership of the Board of Governors shall carry approval, except as otherwise 

Sec. 4. The Board of Governors, when making nominations to office, and to 
the Board, shall endeavor to nominate persons, who in the aggregate are repre- 
sentative of the various branches or organizations of the motion picture industry, 


to the end that there shall be no substantial predominance upon the Board, as the 
result of its own action, of representatives of any one or more branches or organi- 
zations of the industry. 

By-Law IV 


Sec. 1. All committees, except as otherwise specified, shall be appointed by the 

Sec. 2. All committees shall be appointed to act for the term served by the 
officer who shall appoint the committees, unless their appointment is sooner ter- 
minated by the appointing officer. 

Sec. 3. Chairman of the committees shall not be eligible to serve in such ca- 
pacity for more than two consecutive terms. 

Sec. 4. Standing committees of the Society shall be as follows to be appointed 
as designated: 

(a) Appointed by the President and confirmed by the Board of Governors 

Progress Award Committee 

Honorary Membership Committee 

Journal Award Committee 

Admissions Committees 

(Atlantic and Mid-West Sections) 
(Pacific Coast Section) 

European Advisory Committee 
(&) Appointed by the Engineering Vice- President 

Sound Committee 

Standards Committee 

Studio Lighting Committee 

Color Committee 

Theater Engineering Committee 

Exchange Practice Committee 

Non-Theatrical Equipment Committee 

Television Committee 

Laboratory Practice Committee 

(c) Appointed by Editorial Vice- President 

Board of Editors 
Papers Committee 
Progress Committee 
Historical Committee 
Museum Committee 

(d) Appointed by Convention Vice- President 

Publicity Committee 

Convention Arrangements Committee 

Apparatus Exhibit Committee 

(e) Appointed by Financial Vice- President 

Membership and Subscription Committee 

Sec. 5. Two Admissions Committees, one for the Atlantic and Mid- West 
Sections, and one for the Pacific Coast Section, shall be appointed. The former 


committee shall consist of a chairman and six Fellow or Active members of the 
Society of which four shall be members of the Board of Governors. The latter 
committee shall consist of a Chairman and four Fellow or Active members of the 
Society including all officers or members of the Board of Governors of the Society 
residing in the Pacific Coast Section. 

By-Law V 


Sec. 1. The location of each meeting of the Society shall be determined by the 
Board of Governors. 

Sec. 2. Only Honorary members, Fellows, and Active members shall be en- 
titled to vote. 

Sec. 3. A quorum of the Society shall consist in number of one-tenth of the 
total number of Honorary members, Fellows, and Active members as listed in 
the Society's records at the close of the last fiscal year. 

Sec. 4. The fall convention shall be the annual meeting. 

Sec. 5. Special meetings may be called by the president and upon the request 
of any three members of the Board of Governors not including the president. 

Sec. 6. All members of the Society in any grade shall have the privilege of 
discussing technical material presented before the Society or its Sections. 

By-Law VI 

Duties of Officers 

Sec. 1. The president shall preside at all business meetings of the Society and. 
shall perform the duties pertaining to that office. As such he shall be the chief 
executive of the Society, to whom all other officers shall report. 

Sec. 2. In the absence of the president, the officer next in order as listed in 
Article 4 of the Constitution shall preside at meetings and perform the duties of 
the president. 

Sec. 3. The five vice-presidents shall perform the duties separately enumerated 
below for each office, or as defined by the president : 

(a) The executive vice-president shall represent the president in such geo- 
graphical areas of the United States as shall be determined by the Board of Gover- 
nors, and shall be responsible for the supervision of the general affairs of the 
Society in such areas, as directed by the president of the Society. 

(6) The engineering vice-president shall appoint all technical committees. 
He shall be responsible for the general initiation, supervision, and coordination of 
the work in and among these committees. He may act as chairman of any com- 
mittee or otherwise be a member ex-officio. 

(c) The editorial vice-president shall be responsible for the publication of the 
Society's JOURNAL and all other technical publications. He shall pass upon the 
suitability of the material for publication, and shall cause material suitable for 
publication to be solicited as may be needed. He shall appoint a papers com- 
mittee and an editorial committee. He may act as chairman of any committee or 
otherwise be a member ex-officio. 

(d) The financial vice-president shall be responsible for the financial operations 
of the Society, and shall conduct them in accordance with budgets approved by 


the Board of Governors. He shall study the costs of operation and the income 
possibilities to the end that the greatest service may be rendered to the members 
of the Society within the available funds. He shall submit proposed budgets to 
the Board. He shall appoint at his discretion a ways and means committee, a 
membership committee, a commercial advertising committee, and such other 
committees within the scope of his work as may be needed. He may act as chair- 
man of any of these committees or otherwise be a member ex-officio. 

(e) The convention vice-president shall be responsible for the national con- 
ventions of the society. He shall appoint a convention arrangements committee, 
an apparatus exhibit committee, and a publicity committee. He may act as 
chairman of any committee, or otherwise be a member ex-officio. 

Sec. 4. The secretary shall keep a record of all meetings; he shall conduct the 
correspondence relating to his office, and shall have the care and custody of 
records, and the seal of the Society. 

Sec. 5. The treasurer shall have charge of the funds of the Society and disburse 
them as and when authorized by the financial vice-president. He shall make 
an annual report, duly audited, to the Society, and a report at such other times 
as may be requested. He shall be bonded in an amount to be determined by the 
Board of Governors and his bond filed with the Secretary. 

Sec. 6. Each officer of the Society, upon the expiration of his term of office, 
shall transmit to his successor a memorandum outlining the duties and policies 
of his office. 

By-Law VII 


Sec. 1. (a) All officers and five governors shall be elected to their respective 
offices by a majority of ballots cast by the Active, Fellow, and Honorary members 
in the following manner: 

Not less than three months prior to the annual fall convention, the Board of 
Governors shall nominate for each vacancy several suitable candidates. Nomi- 
nations shall first be presented by a Nominating Committee appointed by the 
President, consisting of nine members, including a chairman. The committee 
shall be made up of two Past-Presidents, three members of the Board of Governors 
not up for election, and four other Active, Fellow, or Honorary members, not 
currently officers or Governors of the Society. Nominations shall be made by 
three-quarters affirmative vote of the total Nominating Committee. Such nomi- 
nations shall be final unless any nominee is rejected by a three-quarters vote of 
the Board of Governors present and voting. 

The secretary shall then notify these candidates of their nomination in the 
order of nomination and request their consent to run for office. From the list 
of acceptances, not more than two names for each vacancy shall be selected by 
the Board of Governors and placed on a letter ballot. A blank space shall be 
provided on this letter ballot under each office, in which space the names of any 
Active, Fellow, or Honorary members other than those suggested by the Board 
of Governors may be voted for. The balloting shall then take place. 

The ballot shall be enclosed in a blank envelope which is enclosed in an outer 
envelope bearing the secretary's address and a space for the member's name and 
address. One of these shall be mailed to each Active, Fellow, and Honorary 


member of the Society, not less than forty days in advance of the annual fall con- 

The voter shall then indicate on the ballot one choice for each office, seal the 
ballot in the blank envelope, place this in the envelope addressed to the secretary, 
sign his name and address on the letter, and mail it in accordance with the in- 
structions printed on the ballot. No marks of any kind except those above pre- 
scribed shall be placed upon the ballots or envelopes. 

The sealed envelope shall be delivered by the secretary to a committee of 
tellers appointed by the president at the annual fall convention. This com- 
mittee shall then examine the return envelopes, open and count the ballots, and 
announce the results of the election. 

The newly elected officers and governors of the general Society shall take office 
on the January 1st following their election. 

(b) The first group of vice-presidents, viz., the executive vice-president, 
engineering vice-president, editorial vice-president, financial vice-president, con- 
vention vice-president, and a fifth governor, shall be nominated by the Board of 
Governors at its first meeting after the ratification of the corresponding provisions 
of the Constitution; and the membership shall vote on the candidates in ac- 
cordance with the procedure prescribed in these By-Laws for regular elections of 
officers so far as these may be applicable. 


Dues and Indebtedness 

Sec. 1. The annual dues shall be fifteen dollars ($15) for Fellows and Active 
members and seven dollars and fifty cents ($7.50) for Associate members, pay- 
able on or before January 1st of each year. Current or first year's dues for new 
members, dating from the notification of acceptance in the Society, shall be pro- 
rated on a monthly basis. Five dollars of these dues shall apply for annual sub- 
scription to the JOURNAL. No admission fee will be required for any grade of 

Sec. 2. (a) Transfer a membership may be made effective at any time by pay- 
ment of the pro rata dues for the current year. 

(b) No credit shall be given for annual dues in a membership transfer from a 
higher to a lower grade, and such transfers shall take place on January 1st of each 

(c) The Board of Governors upon their own initiative and without a transfer 
application may elect, by the approval of at least three-fourths of the Board, any 
Associate or Active member for transfer to any higher grade of membership. 

Sec. 3. Annual dues shall be paid in advance. All Honorary members, Fel- 
lows, and Active members in good standing, as defined in Section 5, may vote or 
otherwise participate in the meetings. 

Sec. 4. Members shall be considered delinquent whose annual dues for the 
year remain unpaid on February 1st. The first notice of delinquency shall be 
mailed February 1st. The second notice of delinquency shall be mailed, if neces- 
sary, on March 1st, and shall include a statement that the member's name will be 
removed from the mailing list for the JOURNAL and other publications of the 
Society before the mailing of the April issue of the JOURNAL. Members who are 


in arrears of dues on June 1st, after two notices of such delinquency have been 
mailed to their last address of record, shall be notified their names have been re- 
moved from the mailing list and shall be warned unless remittance is received on or 
before August 1st, their names shall be submitted to the Board of Governors for 
action at the next meeting. Back issues of the JOURNAL shall be sent, if available, 
to members whose dues have been paid prior to August 1st. 

Sec. 5. (a) Members whose dues remain unpaid on October 1st may be 
dropped from the rolls of the Society by majority vote and action of the Board 
or the Board may take such action as it sees fit. 

(6) Anyone who has been dropped from the rolls of the Society for non-pay- 
ment of dues shall, in the event of his application for reinstatement, be considered 
as a new member. 

(c) Any member may be suspended or expelled for cause by a majority vote of 
the entire Board of Governors; provided he shall be given notice and a copy in 
writing of the charges preferred against him, and shall be afforded opportunity 
to be heard ten days prior to such action. 

Sec. 6. The provisions of Sections 1 to 4, inclusive, of this By-Law VIII given 
above may be modified or rescinded by action of the Board of Governors. 

By -Law IX 


Sec. 1. The emblem of the Society shall be a facsimile of a four-hole film-reel 
with the letter S in the upper center opening, and the letters M, P, and E, hi the 
three lower openings, respectively. In the printed emblem, the four-hole openings 
shall be orange, and the letters black, the remainder of the insignia being black 
and white. The Society's emblem may be worn by members only. 

By-Law X 


Sec. 1. Papers read at meetings or submitted at other times, and all material 
of general interest shall be submitted to the editorial board, and those deemed 
worthy of permanent record shall be printed in the JOURNAL. A copy of each issue 
shall be mailed to each member in good standing to his last address of record. 
Extra copies of the JOURNAL shall be printed for general distribution and may be 
obtained from the General Office on payment of a fee fixed by the Board of 

By-Law XI 

Local Sections 

Sec. 1. Sections of the Society may be authorized in any state or locality where 
the Active, Fellow, and Honorary membership exceeds 20. The geographic 
boundaries of each Section shall be determined by the Board of Governors. 

Upon written petition, signed by 20 or more Active members, Fellows and Hon- 
orary members, for the authorization of a Section of the Society, the Board of 
Governors may grant such authorization. 



Sec. 2. All members of the Society of Motion Picture Engineers in good stand- 
ing residing in that portion of any country set apart by the Board of Governors 
tributary to any local Section shall be eligible for membership in that Section, and 
when so enrolled they shall be entitled to all privileges that such local Section may, 
under the General Society's Constitution and By-Laws, provide. 

Any member of the Society in good standing shall be eligible for non-resident 
affiliated membership of any Section under conditions and obligations prescribed 
for the Section. An affiliated member shall receive all notices and publications 
of the Section but he shall not be entitled to vote at Sectional Meetings. 

Sec. 3. Should the enrolled Active, Fellow, and Honorary membership of a 
Section fall below 20, or should the technical quality of the presented papers fall 
below an acceptable level, or the average attendance at meetings not warrant the 
expense of maintaining the organization, the Board of Governors may cancel its 


Sec. 4. The officers of each section shall be a chairman, and a secretary- 
treasurer. The Section chairmen shall automatically become members of the 
Board of Governors of the General Society, and continue in that position for the 
duration of their terms as chairmen of the local sections. Each Section officer 
shall hold office for one year, or until his successor is chosen. 

Board of Managers 

Sec. 5. The Board of Managers shall consist of the Section chairman, the 
Section past-chairman, the Section secretary-treasurer, and six Active, Fellow, or 
Honorary members. Each manager of a Section shall hold office for two years, 
or until his successor is chosen. 


Sec. 6. The officers and managers of a Section shall be Active, Fellow, or 
Honorary members of the General Society. 

Not less than three months prior to the annual Fall Convention of the Society, 
nominations shall be presented to the Board of Managers of the Section by a 
Nominating Committee appointed by the chairman of the Section, consisting of 
seven members, including a chairman. The Committee shall be composed of the 
present chairman, the past-chairman, two other members of the Board of Man- 
agers not up for election, and three other Active, Fellow, or Honorary members of 
the Section not currently officers or managers of the Section. Nominations shall 
be made by a three-quarters affirmative vote of the total Nominating Committee. 
Such nominations shall be final, unless any nominee is rejected by a three-quarters 
vote of the Board of Managers, and in the event of such rejection the Board of 
Managers will make its own nomination. 

The remainder of the procedure shall be in accordance with the procedure 
specified for the election of officers of the General Society as described in By-Law 
VII, Sec. 1A, the word manager being substituted for the word governor. 


Sec. 7. The business of a Section shall be conducted by the Board of Managers. 


Sec. 8. (a) As early as possible in the fiscal year, the secretary of each Section 
shall submit to the Board of Governors of the Society a budget of expenses for the 

(&) The treasurer of the General Society may deposit with each Section secre- 
tary-treasurer a sum of money, the amount to be fixed by the Board of Governors, 
for current expenses. 

(c) The secretary-treasurer of each Section shall send to the treasurer of the 
General Society, quarterly or on demand, an itemized account of all expenditures 
incurred during the preceding interval. 

(d) Expenses other than those enumerated in the budget, as approved by the 
Board of Governors of the General Society, shall not be payable from the general 
funds of the Society without express permission from the Board of Governors. 

(e} A Section Board of Managers shall defray all expenses of the Section not 
provided for by the Board of Governors, from funds raised locally by donation, 
or fixed annual dues, or by both. 

(/) The secretary of the Society shall, unless otherwise arranged, supply to 
each Section all stationery and printing necessary for the conduct of its business. 


Sec. 9. The regular meetings of a Section shall be held in such places and at such 
hours as the Board of Managers may designate. 

The secretary-treasurer of each Section shall forward to the secretary of the 
General Society, not later than five days after a meeting of a Section, a statement 
of the attendance and of the business transacted. 


Sec. 10. Papers shall be approved by the Section's papers committee previ- 
ously to their being presented before a Section. Manuscripts of papers presented 
before a Section, together with a report of the discussions and the proceedings of 
the Section meetings, shall be forwarded promptly by the Section secretary- 
treasurer to the secretary of the General Society, Such material may, at the dis- 
cretion of the Board of Editors of the General Society, be printed in the Society's 

Constitution and By-Laws 

Sec. 11. Sections shall abide by the Constitution and By-Laws of the Society 
and conform to the regulations of the Board of Governors. The conduct of Sec- 
tions shall always be in conformity with the general policy of the Society as fixed 
by the Board of Governors. 


By-Law XII 


Sec. 1. These By-Laws may be amended at any regular meeting of the Society 
by the affirmative vote of two-thirds of the members present at a meeting who are 
eligible to vote thereon, a quorum being present, either on the recommendation of 
the Board of Governors or by a recommendation to the Board of Governors signed 
by any ten members of active or higher grade, provided that the proposed amend- 
ment or amendments shall have been published in the JOURNAL of the Society, 
in the issue next preceding the date of the stated business meeting of the Society 
at which the amendment or amendments are to be acted upon. 

Sec. 2. In the event that no quorum of the voting members is present at the 
time of the meeting referred to in Section 1, the amendment or amendments shall 
be referred for action to the Board of Governors. The proposed amendment or 
amendments then become a part of the By-Laws upon receiving the affirmative 
vote of three-quarters of the Board of Governors. 


Prior to January, 1930, the Transactions of the Society were published quar- 
terly. A limited number of these Transactions are still available and will be 
sold at the prices listed below. Those who wish to avail themselves of the op- 
portunity of acquiring these back numbers should do so quickly, as the supply 
will soon be exhausted, especially of the earlier numbers. It will be impossible 
to secure them later on as they will not be reprinted. 























Beginning with the January, 1930, issue, the JOURNAL of the Society has been 
issued monthly, in two volumes per year, of six issues each. Back numbers of 
all issues are available at the price of $1.00 each, a complete yearly issue totalling 
$12.00. Single copies of the current issue may be obtained for $1.00 each. 
Orders for back numbers of Transactions and JOURNALS should be placed through 
the General Office of the Society and should be accompanied by check or money- 


The following are available from the General Office of the Society, at the prices 
noted. Orders should be accompanied by remittances. 

Aims and Accomplishments. An index of the Transactions from October, 
1916, to December, 1929, containing summaries of all articles, and author and 
classified indexes. One dollar each. 

Journal Index. An index of the JOURNAL from January, 1930, to December, 
1935, containing author and classified indexes. One dollar each. 

Motion Picture Standards. Reprints of the American Standards and Recom- 
mended Practice as published in the March, 1941, issue of the JOURNAL; 50 cents 

Membership Certificates. Engrossed, for framing, containing member's name, 
grade of membership, and date of admission. One dollar each. 

Journal Binders. Black fabrikoid binders, lettered in gold, holding a year's 
issue of the JOURNAL. Two dollars each. Member's name and the volume 
number lettered in gold upon the backbone at an additional charge of fifty cents 

Test-Films. See advertisement in this issue of the JOURNAL. 




Volume XXXVI May, 1941 



Reduction of Development Sprocket- Hole Modulation 


A New Mirror Light-Modulator W. R. GOEHNER 488 

Scanning Theory S. SABAROFF 497 

Stability of Synchronous Motors 


Note on the Keeping of Hydrogen Peroxide-Ammonia Hypo 

Eliminator Solutions 


Report on Arc Lamp Noise Tests, from the Research Council, 
Academy of Motion Picture Arts and Sciences 559 

Current Literature 572 

Rochester Convention, Hotel Sagamore, May 5th-8th, Inclusive 573 

Abstracts of Convention Papers 577 

Society Announcements 586 





A. C. DOWNES, Chairman 




Subscription to non-members, $8.00 per annum; to members, $5.00 per annum, 
included in their annual membership dues; single copies, $1.00. A discount 
on subscription or single copies of 15 per cent is allowed to accredited agencies. 
Order from the Society of Motion Picture Engineers, Inc., 20th and Northampton 
Sts., Easton, Pa., or Hotel Pennsylvania, New York, N. Y. 

Published monthly at Easton, Pa., by the Society of Motion Picture Engineers. 

Publication Office, 20th & Northampton Sts., Easton, Pa. 

General and Editorial Office, Hotel Pennsylvania, New York, N. Y. 

West Coast Office, Suite 226, Equitable Bldg., Hollywood, Calif. 

Entered as second class matter January 15, 1930, at the Post Office at Easton, 
Pa., under the Act of March 3, 1879. Copyrighted, 1941, by the Society of 
Motion Picture Engineers, Inc. 


** President: EMERY HUSE, 6706 Santa Monica Blvd., Hollywood, Calif. 
** 'Past-President: E. ALLAN WILLIFORD, 30 East 42nd St., New York, N. Y. 
** Executive Vice-P resident: HERBERT GRIFFIN, 90 Gold St., New York, N. Y. 

* Engineering Vice-President: D. E. HYNDMAN, 350 Madison Ave., New York. 
**Editorial Vice-President: ARTHUR C. DOWNES, Box 6087, Cleveland, Ohio. 

*Financial Vice-President: A. S. DICKINSON, 28 W. 44th St., New York, N. Y. 
** Convention Vice-President: W. C. KUNZMANN, Box 6087, Cleveland, Ohio. 

*Secretary: PAUL J. LARSEN, 44 Beverly Rd., Summit, N. J. 

*Treasurer: GEORGE FRIEDL, JR., 90 Gold St., New York, N. Y. 


**M. C. BATSEL, 501 N. LaSalle St., Indianapolis, Ind. 

*J. A. DUBRAY, 1801 Larchmont Ave., Chicago, 111. 

"JOHN G. FRAYNE, 6601 Romaine St., Hollywood, Calif. 

*A. N. GOLDSMITH, 580 Fifth Ave., New York, N. Y. 

*ARTHUR C. HARDY, Massachusetts Institute of Technology, Cambridge, Mass. 
**L. L. RYDER, 5451 Marathon St., Hollywood, Calif. . 

*T. E. SHEA, 195 Broadway, New York, N. Y. 

*R. O. STROCK, 35-11 35th St., Astoria, L. I., N. Y. 

*Term expires December 31, 1941. 
**Term expires December 31, 1942. 



Summary. One of the contributing factors to sound-track degradation is sprocket- 
hole modulation. This is more commonly known as 96-cycle modulation, and results 
from non-uniform action of developer around the perforations during the time of 
processing. Its chief remedy is turbulation. 

The practical aspects of controlling the amount of sprocket-hole modulation is 
described herein. Curves showing the increase of this distortion due to diminished 
turbulation are included as well as those showing the inter modulation of recorded 
sound by sprocket-hole agitation. The turbulation method employed at the Film 
Laboratory of Twentieth Century- Fox Film Corporation at Hollywood is disclosed 
and the various sensitometric means of control relative to this problem are given. 

One of the many problems confronting the laboratory and sound 
technicians of the motion picture industry is the elimination of 96- 
cycle sprocket-hole modulation. The use of noise-reduction systems 
and the improvement in other aspects of recording and processing 
have made this problem daily more obvious and acute. 

There are many and varied causes of sprocket-hole modulation. 
In general, they are : (1) variation in film-operating speed at sprocket- 
hole frequency in recording, printing, and projection; (2) variation 
in exposure during these operations due to reflections, etc.; and (3) 
variation of development around the sprocket-hole areas. Special 
cases due to abrasions and pressure occur now and then, but their 
causes and methods of prevention are well known. 

The improvements in film-drive mechanisms in the recording ma- 
chines and the adoption of the new type sound-heads in projection 
have reduced the 96-cycle modulation due to variation of film speed 
and exposure in recording and projection to the point where the major 
troubles remain in processing and printing. It is the purpose of this 
paper to show how, at the Twentieth Century-Fox Laboratories, 

*Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received October 
15, 1940. 

"Twentieth Century-Fox Film Corp., Hollwyood, Calif. 


$> The Society is not responsible for statements by authors <> 

470 LESHING, INGMAN, AND PIER [j. s. M. P. E. 

sprocket-hole modulation resulting from development has been re- 
duced to the point at which the 96-cycle component is inaudible in 
the ground-noise. 

The first goal in attacking the problem was to find the simplest 
and most practicable method of measuring the results of our experi- 
ments. Frayne 1 had used successfully at least three methods in 
his investigation of the effect of the sprocket-holes upon development 
of the sound-track area, and in his comparisons of the processing 
conditions obtaining at the different film laboratories in Hollywood. 
With a flutter-measuring set he measured the frequency and ampli- 
tude modulation of a 3000-cycle wave, and by scanning the sound- 
track in 5-mil increments showed that the sprocket-holes produced a 
marked effect upon the development of the track over a distance of 
approximately 30 mils from the holes. Another method used was 
that of scanning an unmodulated track with a microphotometer, 
which gave substantially the same results. A further method suc- 
cessfully used was the scanning, by a microphotometer, of an exposure 
of square-wave characteristics. This test illustrated the advantages 
of some systems of turbulation of the developer. 

The time-honored test of developing sensitometer strips in opposite 
directions was considered at Twentieth Century-Fox but was dropped 
as not being sufficiently critical. Comparison of equal exposures of 
different areas was tried and found to be of value in estimating the 
effectiveness of turbulation of the developer. 

The most common testing methods employed were the measure- 
ment of the 96-cycle tone in an unmodulated track by running the 
track on a sound-head and measuring the signal through a band-pass 
filter with a volume indicator, and by comparing the densities and 
gammas as measured between the sprocket-holes and in the sound- 
track area. The latter test proved to be the most critical of all, and, 
as the measurements could be made in the film laboratory on existing 
equipment, was used to a great extent. 

In describing these tests one might be led to think that some of the 
tests have no connection with sprocket-holes or sound-track areas. 
It must be remembered that directional effect and sprocket-hole 
modulation due to development are caused by the same thing and 
can be prevented by the same methods. Failure to remove the 
restraining products of development, that is, the soluble bromides, as 
fast as they reach the surface of the gelatin is the cause of these de- 
fects. Naturally, therefore, if it were possible to remove these 

May, 1941] 



products before they could affect adjacent areas, there would be none 
of the aforementioned effects present. These effects are most serious 
in cases where a high gamma-infinity stock, such as sound-recording 
film, is developed to a low gamma. 

It was evident that if the length of time that the exhausted developer 
was in contact with the film were lessened sufficiently, there would be 
no change in the result of development from the additional agitation 
of the sprocket-holes or from the directional effect. If the flow of 
developer on the surface of the emulsion could take place at a rate 
high enough to wash away the bromides and supply 
unexhausted developer before any deleterious effects 
were produced, then the ideal condition would have 
been reached. 

In order to determine the rate of flow necessary 
to approximate this condition, a special developing 
machine was built, on which short lengths of film 
were developed. The speed of the drum carrying the 
film was controlled very definitely, so that the rate 
of flow of the developing solution over the surface 
of the emulsion, which ranged from zero to 1750 feet 
per minute, could be measured. 

Exposures were made on the Eastman lib sensi- 
tometer so that they could be read between the 
sprocket-holes as well as in the sound-track position 
(Fig. 1). The small dots are of identical exposure, 
and are of value when compared with the larger ex- 
posures in estimating the effectiveness of the removal 
of development products. The increase in density 
should be noted of the areas between the sprocket- 
holes due to the increased agitation of the developer. This film was 
travelling through the developer at a rate of 90 feet per minute and was 
processed under the conditions prevailing before the addition of the 
present system of turbulation. 

Fig. 2 shows the relation between densities measured between the 
holes (broken curve) and in the sound-track area (solid curve), for 
the strip shown in Fig. 1. The separation between the curves is a 
direct measure of the amount of sprocket-hole modulation to be 
expected, inasmuch as it is this density difference that produces the 
96-cycle modulation. 

As the speed of the film through the developer is increased it is 

FIG. 1. Pho- 
tograph of 
special sensi- 
tometer strip. 



[J. S. M. P. E 

to be expected that the more complete development of the exposures 
due to increased agitation will result in greater density and contrast. 
Fig. 3 shows this to be a fact. The time of development was held 
constant and the speed of the film through the developer was varied 
from zero to 1750 feet per minute. The increase of gamma in the 
sound-track is shown by the solid curve and between the sprocket- 
holes by the broken curve. The fact that the broken curve begins to 
flatten at a lower film speed than the solid curve is evidence of a 
closer approach to ideal conditions, and is a result of increased agita- 
tion of the developer in the sprocket-hole areas. 


FIG. 2. Relation between densities measured be- 
tween sprocket-holes (broken curve) and in the sound- 
track area (solid curve), for the strip shown in Fig. 1; 
developed at 90 fpm ; no turbulation. 

Fig. 4 shows the increase in density with increase in speed through 
the developer. The general characteristics of these curves are the 
same as those of Fig. 3. Note that above 700 feet per minute further 
increases of speed produce no appreciable difference in density be- 
tween the holes. The exposure used for this measurement was the 
llth step, which gives practically the same exposure as is required 
for the unbiased sound-track. The vertical distance between the 
curves is a measure of the 96-cycle modulation produced in develop- 

The difference in density between the sound-track area and the 
areas between the sprocket-holes is shown by Fig. 5. The tests at 

May, 1941] 



various film speeds were developed to the same gammas. As can be 
seen, the maximum amount of sprocket-hole modulation occurs at 
speeds between 25 and 100 feet per minute. The worst conditions 
prevail at the very speeds at which the developing machines were 
designed to operate. 

Many schemes were considered for increasing the agitation of the 
developer. The use of a trough with the developer running at high 
speed opposite to the film travel has been tried with some success 



FIG. 3. Curves showing difference between gamma of areas between 
sprocket-holes (broken curve) and sound-track area (solid curve), as a func- 
tion of film speed; developed five minutes; no turbulation. 

but would have necessitated building new developing machines. 
Squeegees and brushes in contact with the emulsion have been used 
effectively in some "still" laboratories and might have value in con- 
junction with some types of developing machines. At Twentieth 
Century-Fox the Spoor-Thompson machine is used, in which the film 
runs vertically in long strands. Between the upper and lower rollers 
there is no support for the film and therefore no pressure may be ap- 
plied to the film except possibly at the rollers, which would represent 
a very small part of the total footage in the developer and would not 
be very effective. Increasing the speed of the machine would be im- 



[J. S. M. P. E. 

practicable because of the risk of film breakage, which becomes great 
at speeds over 125 feet per minute. As mechanical agitation and 
greater machine speed were impracticable, we decided to try increas- 
ing the relative speed of the film by the use of some system of moving 
the developer across the film at a rapid rate. Spraying the film with 
developer forced through jets at considerable pressure probably would 
be the best way of removing the development products from the 
surface of the emulsion, but the problem of supporting the film 
against this pressure precluded the use of such a method. 



FIG. 4. Difference between density of areas between sprocket-holes 
(broken curve) and sound-track area (solid curve), as a function of speed 
of film; developed five minutes; no turbulation. 

It was finally decided to try emptying the developing tanks and 
flowing the developer down the vertical strands of film. Header jets 
at low pressure were used for the purpose, allowing the developer to 
flow by gravity down the full length of the strand. It was realized 
that the relative speed of the developer over the film would not be 
sufficient to achieve the ideal state of turbulation under which there 
would be no difference between the densities around the sprocket- 
holes as compared with those in other areas. In order to minimize the 
directional effect, the developer was caused to flow down the film 

May, 1941] 



in the direction of film travel as well as in the opposite direction. The 
rate of flow of developer through the jets was 250 gallons per minute 
per machine. Fig. 6 shows the arrangement of the jets and the flow 
of developer down the film. All our developing machines are now 
operating substantially as shown and described. Actually, the tanks 
are not completely empty. Sufficient developer is kept in the bottom 
of the tanks to provide lubrication for the lower rollers. 

The results obtained with this method of development are sur- 
prisingly good. Fig. 7 represents the conditions as they exist using 


FIG. 5. Difference in density between sprocket-hole areas and sound-track 
area, as a function of film speed; developed to gamma 0.40. 

the machines so modified. Note the small separation between the 
two curves as compared with the curves of Fig. 2, for the machine in 
its previous condition. There is still some difference between the 
curves but the measurements made between the holes represent the 
most extreme condition of turbulation to be found on the surface of 
the film. The difference in density at the llth step now amounts to 
only 0.04, whereas with the old system the difference at that step 
was 0.18 (Fig. 2). Returning to Fig. 5, the differential of 0.04 that 
we now have is seen to be equivalent to the results obtained by 

482 LESHING, INGMAN, AND PIER [j. s. M. P. E. 

running the developing machine at a speed of 400 feet per minute. 
This result is now attained at a safe machine speed of only 90 to 100 
feet per minute. 

Electrical Research Products, Inc., in their investigations of process- 
ing conditions at the various laboratories in Hollywood have found 
that the Twentieth Century-Fox Laboratory is consistently good with 
respect to the negligible 96-cycle modulation present in the product. 
One of their tests is particularly interesting from the picture point of 

FIG. 6. Photograph showing the arrangement of the jets and the flow of 
developer on the surface of the film. 

view as well as the sound. This is the square-wave exposure referred 
to earlier in this paper. Fig. 8 was produced by scanning the block 
exposures with a microphotometer in a direction parallel to that in 
which the film was developed. The figure on the left was obtained 
from the tests processed at Twentieth Century-Fox Laboratories 
while that on the right was developed at the laboratories of another 
major studio. The right-hand figure exhibits very marked directional 
effects. Observe how the entering edge of the exposure is developed 
quite efficiently, and how the development of the rest of the block is 
restrained to a greater and greater degree as the soluble bromides are 

May, 1941] 



dragged across the exposure. The sample developed by us shows 
practically no directional effect and considerably less Eberhard ef- 
fect. The fact that the density at the edges of the exposure is closer 
to the density in the center of the exposure, and the underdeveloped 
areas adjacent to the high exposures are of less magnitude, makes it 
seem reasonable to assume that there is some reduction of the Eber- 
hard effect. This assumption requires further investigation before 
it can be proved, and it is contrary to the expressed opinion of some 


L.Q6 E 



FIG. 7. Curves showing difference between density of areas 
between sprocket-holes (broken curve) and sound-track area (solid 
curve), developed at 90 fpm; with turbulation. (Compare with 
Fig. 2.) 

experts in this line. It is our intention to continue investigations in 
this direction. 

In order that the results of these tests may be comparable with 
directional tests made by other investigators, we present here the 
same type of test as used by J. Crabtree and J. H. Waddell 2 and 
others (Fig. 9) . Sensitometer strips were developed in opposite 
directions in the developing machine under the present conditions 
and also under the original conditions. The machine speed in both 
cases was 90 feet per minute. The broken curve lines represent 
development with the high exposures leading. The negligible di- 



LJ. S. M. p. E. 

rectional effect shown by the curves on the right, representing the 
conditions as they are at present is apparent. 

In comparative measurements on a sound-head through a low-pass 
filter we find that the negative reproduces about 6 to 10 db less of 
the 96-cycle modulation with our present system than previously. 
This reduction is sufficient to eliminate development sprocket-hole 
modulation completely from the print so far as the ear is concerned. 
Only by eliminating ground-noise by the use of band-pass filters is it 
possible to measure the modulation now present. 









,_1 INCH , 

>_1 INCHJ 









FIG. 8. Square-wave test: (left) From tests processed at Twentieth 
Century-Fox Laboratories, (right) from tests developed at laboratories of 
another major studio. 

Many interested persons have asked about the comparative chemi- 
cal costs with this method of turbulation. At first glance it would 
appear that, with the developer flowing through the air, the oxidation 
of the developing agents would be serious and that the resulting loss 
of developing power would require that larger quantities of the solu- 
tions be thrown away and more booster used, thus increasing the 
chemical costs materially. On the contrary, we found that the 
greatly increased efficiency of the developer, due to the more effective 
agitation, forced us to cut the quantity of developing agents to less 

May, 1941] 



than half the original amount. Another factor is that aeration of a 
developer of low alkalinity containing hydroquinone has been found 
to increase the developing power to some extent. The rate at which 
used developer is discarded is about the same as before. A reasonable 
increase in the cost of processing would not have deterred us from 
using any methods resulting in an improvement in our product, but 
it was with some sense of satisfaction that this particular advance- 






FIG. 9. Comparison of tests made with and without turbulation: (broken 
curves) strips developed with high exposures leading; (solid curves) strips 
developed with low exposures leading. 

ment was achieved at no additional operating expense. As a matter 
of fact, the only difficulties which we have experienced are attributable 
to the small amount of chemicals in the solution. It is necessary to 
exercise somewhat closer control of time of development and addi- 
tion of booster than was customary previously. 

Although this paper primarily pertains to sound-track develop- 
ment, it will be obvious that the superior turbulation that resulted in 
the reduction of sprocket-hole modulation development will be ef- 
fective in alleviating directional effects, Mackie lines, etc., in the 
picture as well. Undoubtedly there is much room for further im- 


provement of developing machines, as all machines at present fail 
to achieve ideal turbulation by a large margin. We are conscious of 
the fact that this is a rather makeshift arrangement and that by 
building an entirely new developing machine still better results could 
be achieved. As an addition to the Spoor-Thompson machine, how- 
ever, we feel that the benefits derived are great enough to make it 
worthy of being brought to the attention of the Society. 


1 FRAYNE, J. G., AND PAGLIARULO, V.: "The Influence of Sprocket-Holes 
upon the Development of Adjacent Sound-Track Areas." /. Soc. Mot. Pict. 
Eng., XXVIII (March, 1937), p. 235. 

2 CRABTREE, J., AND WADDELL, J. H.: "Directional Effects in Sound-Film 
Processing," J. Soc. Mot,. Pict. Eng., XXI (Nov., 1933), p. 351. 


MR. CRABTREE : Mr. Leshing and his co-workers are to be congratulated on 
this ingenious and simple method of increasing turbulation. It is an application 
of the scheme used by Friend Baker at the Warner Studios many years ago, the 
film being developed on a reel. The developer was poured over the reel across 
the surface of the film and the developer turned the reel in the manner of a water- 

MR. DAILY: Is the sound negative developer agitation .method applied also to 
the release positive developing machines? 

MR. LESHING: We are not making any release prints here to any extent. We 
make only five or seven copies so far as those are concerned, yes. I do not 
know exactly what method is being used in the DeLuxe Laboratories in the 

MR. KELLOGG: This method of agitation permits greater motion of liquid 
over the sound-track than it does over the region between the sprocket-holes, be- 
cause the sprocket-hole, when the flow is parallel to the film, is somewhat of an 

The question occurs as to whether there might not be too much oxidation of the 
developer, with it continually flowing over the surface of the film. 

MR. CRABTREE : About six months ago we published a paper in the JOURNAL 
dealing with the effect of excessive aeration of developers, and the types used 
were surprisingly resistant to aerial oxidation. 

In the case of developers containing hydroquinone, sodium hydroxide is formed 
as a by-product of oxidation, so that the activity of the developer increases. In 
the case of developers containing metol, there is no such effect, but if the activity 
of the developer falls off it is simply a matter of adjusting the rate of replenish- 
ment to compensate for any slight change in activity. 

MR. KELLOGG: I am familiar with the method of agitation by sending air 
bubbles through the developer. But allowing the developer to flow down the 
film in thin surface streams might expose it to the air a great deal more than the 
bubble system. 


I would be interested in knowing whether it is a question of degree of oxidation, 
or whether the developer simply refuses to be oxidized any more due to some kind 
of exhaustion or saturation effect. 

MR. LESHING: We have never made any measurements of the rate of oxida- 
tion. However, any fears that we may have had that the developer would oxidize 
or would not stand up, or that we should have to use extremely strong formulas, 
were dispelled with the first test. We tried all kinds of jets flat, round, and of 
different sizes but found nothing to indicate that the rate of oxidation is greater 
with one form of turbulation than with any other. 

MR. CRABTREE: Do you have any difficulty due to excessive foaming of the 
developer, or do you use an anti-foaming agent? 

MR. LESHING : We use some ; not much. 

MR. SCOVILLE: This matter of turbulation goes much further than merely 
improving the 96-cycle modulation, and affects to a marked extent the choice of 
optimum processing conditions for variable-density sound. 

As an example, if dynamic recording tests are made and the film is developed 
at Fox, the optimum print and negative densities determined thereby are very 
nearly the same as would be expected from the sensitometric data. Also, there 
is good correlation between the indications of various types of distortion tests, 
such as delta-db tests, intermodulation tests, and harmonic tests. With film de- 
veloped at other Hollywood laboratories no such correlation is generally ob- 

The question was brought up a while ago as to why a dynamic recording test 
does not indicate the same optimum as the sensitometer test. Such investiga- 
tions as I have made indicate that directional effect and lack of adequate turbula- 
tion are usually responsible for that difference. 

DR. FRAYNE: In our tests of sound-track development we have always found 
in the Fox Laboratories less 96-cycle modulation, less directional effect, and also 
the best correlation between sensitometric control and other tests such as Mr. 
Scoville has pointed out. 

I hope that this paper will stimulate other laboratories to do as good a job in 
this regard as Mr. Leshing has accomplished. 



Summary. A vibrating mirror light-modulator developed for 200-mil push" 
pull recording is described. Modern magnetic materials are used in the magnetic 
structure to obtain high electromagnetic efficiency and therefore adequate electro- 
magnetic damping. Distortion is limited to low values by accurate mechanical bal- 
ance of the armature. An inverse shunting network, to build the modulator impe- 
dance out to a constant resistance, has been developed, making the response characteris- 
tic substantially free from driving-circuit impedance, so that any desired frequency 
characteristic can be obtained by the insertion of a constant-resistance equalizer. 

The technic of recording sound on film involves a series of con- 
versions from one form of energy to another. The first step is the 
conversion of the acoustic energy of the sounds to be recorded into 
electrical energy. The final step, a chemical one, produces a metallic 
silver strip which should be, if the complete process is correct, an 
exact record of the instantaneous variations of the original acoustic 
energy. Since there are several stages in this conversion process, 
the possibility exists at each stage that the new form of energy will 
not be a true copy of the original. Much of the development work 
that has been carried on since the introduction of the sound-on-film 
method for commercial sound pictures has been directed toward 
the objective of reducing to very low values any remaining distortion 
in each step of the process. Comparison of present-day recording 
quality with some of the early recordings gives ample evidence of 
the progress that has been made in the entire process from micro- 
phone to the final image on the film, and from film to the re-created 
sound in the theater. The present state of the art, therefore, im- 
poses quite severe requirements on new equipment, especially when 
further improvements in overall quality are desired. Consideration 
of these requirements led Electrical Research Products engineers to 
the use of 200-mil push-pull variable-area sound-track for prime 

* Presented at the 1940 Meeting at Hollywood, Calif. ; received Oct. 10, 

** Bell Telephone Laboratories, New York, N. Y. 

^ The Society is not responsible for statements by authors $ 



recordings as a direct means of improving the overall sound quality 
of the release print. To accomplish this objective, it became evident 
that a new recording system, including an improved light-modulator, 
would be required if the full advantage of the proposed recording 
method were to be realized. This paper describes the vibrating- 
mirror modulator developed for this system by the Bell Telephone 

The optical system for push-pull recording, as shown by Fig. 1, 
was designed to use a mirror 6.7 X 10.0 mm. With this arrange- 
ment, a deflection of approximately ==0.3 degree will modulate both 
tracks to the 100 per cent value. It is evident that a smaller mirror 
placed closer to the recording slit can be adjusted to pass the required 
light flux, but greater amplitude will be necessary to produce 100 
per cent modulation. 


FIG. 1. Optical system. 

In the selection of an electromechanical driving system that 
would rotate the rather large mirror through the required angular 
distance, the Laboratories' experience in other fields was drawn on 
rather heavily. Although the Laboratories had not previously 
developed a vibrating-mirror light-modulator for use in film record- 
ing, it was found that many devices developed for other purposes, 
such as galvanometers, oscillographs, relays, and receivers had fea- 
tures that could be directly applied to this light-modulator problem. 
N Practical experience with some of these structures influenced the 
selection of a mechanical coupling to the mirror that would eliminate 
the possibility of lost motion between the applied torque and the 
motion of the mirror. The necessity for this precaution is evident 
when it is realized that the amplitude of the end of the mirror at 100 
per cent modulation is only l / m of the 6.7-mm dimension, or ap- 
proximately 0.0007 inch. After some preliminary tests, a balanced 



[J. S. M. P. E. 

magnetic structure, 1 basically capable of design for very low mag- 
netic distortion, was selected as offering the best possibility of meet- 
ing the mechanical and electrical distortion limits. 

A cross-section of the magnetic structure is shown in Fig. 2. The 
mirror is mounted directly on the removable armature, shown in 
Fig. 3. The torque to drive the mirror is obtained along the edge 
of the armature by the magnetic pull over the pole-tips. The pole- 
pieces are built up from laminations of the shape shown in Fig. 3. 

FIG. 2. Magnetic structure with armature removed. 

These laminations are securely clamped by two grooved bars, one 
of which is shown in Fig. 2. The pole-piece assembly is held against 
the permanent magnet by the non-magnetic centering clamps. The 
steel screws holding these clamps pass through clearance holes, per- 
mitting accurate centering of the pole-pieces with respect to the 
bridge. The armature, with its torsion supports and frame, is milled 
from a single piece of metal; it rests on spacers and the tensioned 
tungsten wire shown in Fig. 3. The tungsten wire serves the two- 
fold purpose of accurately determining the air-gap and at the same 
time providing a precision, line-contact bearing surface of extreme 

May, 1941] 



hardness. Both sides of the armature are lapped and polished 
to provide a satisfactory mounting for the mirror. The coils are 
layer-wound, permitting the coils to fit into the small space left 
between the bridge and the edge of the pole-pieces. As shown in 
Figs. 2 and 3, the path of the steady flux from the permanent magnet 
is through the pole-tips to both sides of the armature, passing through 
the armature to the center above the bridge, leaving the armature 
above the tungsten wire to the bridge, and out both sides of the 
bridge, returning to the base of the permanent magnet through the 
{/-shaped supporting structure. The signal windings placed around 
the pole-pieces are connected so that the flux due to the signal cur- 
rents aids the steady flux in one gap and opposes it in the other, 



FIG. 3. Pole-piece and armature. 

thereby causing the mirror to oscillate in accordance with the direc- 
tion and magnitude of the signal current. Since one side of the 
armature approaches the pole-piece when the other side of the 
armature is receding, the total air-gap remains constant. With 
this type of structure, the second-harmonic distortion is balanced 
out if the mechanical structure is balanced ; for this reason consider- 
able care was taken to make the structure symmetrical about the 
rotational axis. Since the mechanical balance may not be perfect, 
the remaining distortion can be minimized by designing the magnetic 
circuit to reduce the initial value of the distortion caused by each 

The design of the magnetic circuit to minimize the magnetic dis- 
tortion is very similar to the design of vacuum tube amplifier cir- 
cuits, where the non-linear relation between plate current and grid 
voltage can be made to yield, for a limited voltage range, any degree 



[J. S. M. P. E. 

of linearity, depending upon what elements are connected in the 
circuit. In amplifier design experience, it has been found that less 
than 1 per cent distortion can not be heard under optimum condi- 
tions of listening with wide -frequency band systems. If the record- 
ing-reproducing process were developed to the ideal limit, it might 



6 8 10 12 14 16 18 20 


22 24 

XIO 3 

FIG. 4. Reversible permeability vs. polarizing flux density. 

be desirable to hold the distortion in light-modulators to this value 
or less. However, film as a recording medium has some non-linear 
distortion by itself, as evidenced by cross-modulation and develop- 
ment effects. For this reason, it does not seem profitable at the 
present time to reduce the modulator distortion to less than the over- 
all film development effects. This modulator was designed to limit 
the harmonic distortion to 2 per cent for full modulation. At lower 
modulations, the relative level of the harmonics decreases rapidly 

May, 1941] 



to less than the 1 per cent value. If lower harmonic levels are re- 
quired by future improvements in the film characteristics, the same 
design principles can be applied to meet the new limits. 

In addition to the requirements of amplitude, mirror size, and low 
harmonic distortion, three other objectives were established: (1) 
the frequency response should be flat to 8000 cycles, (2) the damp- 
ing should be adequate to prevent noticeable "overhang," and (3) 
the sensitivity should be high enough to permit operation from exist- 
ing amplifiers. These three requirements are interrelated and 
directly dependent upon the electromechanical efficiency of the 

200 500 1000 2000 


FIG. 5. Impedance of modulator. 


torsional motor. It can be shown that in the ideal case the torque 
developed by this type of motor is 

N A W B , 

A = Area of pole-piece 
W = Width of armature 
B Q Polarizing flux 
d = Air-gap 
N = Turns on coil 
/ = Current 

This equation is based upon the assumption that air-gap reluctance 
is very large compared to the iron-path reluctance and that all the 
flux due to the current / links the air-gap. If these conditions ap- 



[J. S. M. P. E. 

plied to a practical design, it would be desirable to make B as large 
as possible and the air-gap as small as is practicable. As might be 
expected, the ideal case does not apply fully, the difficulty being 
that the reluctance of the iron circuit can not be made small enough 


o V\A/ p^5^-t 

FIG. 6. Approximate equivalent circuit of 

and, further, that the reversible permeability of the magnetic ma- 
terial is a function of the superposed steady flux. The effect of this 
factor upon the efficiency of polarized motor structures has been 
shown to reach a maximum at a value of superposed steady flux 
that depends upon the choice of magnetic design and magnetic ma- 
terials. 2 These data are shown in Fig. 4, which shows that the mod- 
ern magnetic materials are su- 
perior in regard to the magni- 
tude of reversible permeability 
in the presence of high polariz- 
ing flux densities. In particular, 
both permendur and 45 perm- 
alloy are shown to be superior 
to the older magnetic materials. 
Permendur, offering the highest 
force factor and, therefore, the 
highest electromagnetic damping 
and efficiency, was selected for 
use in the parts of the circuit 
carrying high flux density, i. e., the armature, pole-pieces, and bridge. 
The impedance of the modulator, as shown in Fig. 5, can be 
represented approximately by the circuit arrangement shown in 
Fig. 6; the values of the circuit elements can be determined from the 
impedance data. 3 Once the values of the equivalent network are 
determined, a shunting network of a type developed to equalize a 
telephone receiver can be designed to make the load impedance 


FIG. 7. Configuration of constant- 
impedance network. 

May, 1941] 



looking into the modulator constant. 4 -* Fig. 7 shows the configura- 
tion of this network. Fig. 8 shows the impedance of the modulator 
when connected to a constant-impedance network of the type shown 
in Fig. 7. The maximum impedance variations are reduced to the 


o w o 
o o o o c 




200 500 1000 2000 


5000 10,000 

FIG. 8. Modulator impedance with constant-resistance network. 

order of 4 per cent and the average, over the major part of the fre- 
quency range, is approximately == 1 per cent. This means that the 
response characteristic of the modulator is made substantially inde- 
pendent of the driving-circuit impedance and that any desired re- 
sponse can be predicted and obtained by the insertion of an equalizer 
of the constant-resistance type. 

FIG. 9. Variable-area modulator. 

In addition to the electrical, mechanical, and magnetic problems, 
considerable attention was given to the optical quality of the mirror 
surface. The mirror surface was worked to the flatness limits re- 
quired by the optical system. To insure maintenance of the optical 
flatness of the mirror when mounted on the armature, a special 

* Modulator network designed by E. L. Norton. 

496 W. R. GOEHNER 

cementing technic was developed. Overall distortion of less than one 
wavelength (0.00002 inch) from a true surface has been found prac- 
ticable. The mirror surface is aluminum, evaporated onto the 
highly polished front surface of the glass, providing a mirror of high 
reflecting power well into the ultraviolet region. 

The overall dimensions of the modulator mounted in its case are 
! 3 /4 X 1 3 A X 2 Vie inches. Fig. 9 is a photograph of the assembled 


1 HARRISON, H. C.: U. S. Pat. applied for. 

2 LEGG, V. E.: "Survey of Magnetic Materials and Applications in Telephone 
Systems," Bell Syst. Tech. J. (July, 1939), p. 438. 

3 WEGEL, R. L.: "Theory of Magneto-Mechanical Systems as Applied to 
Telephone Receivers and Similar Structures," J. A. I. E. E., XL (Oct., 1921), p. 

4 JOHNSON, K. S. : "Transmission Circuit for Telephonic Communication," 
D. Van Nostrand & Co. (1927), p. 237. 


Summary. Scanning, or probing, is fundamental to most objective investigation. 
A spatial scanning theory applicable to a space of any number of dimensions is de- 
scribed in which the relation bet-ween the field being scanned and the scanned output is 
expressed in terms of a scanning operator, or function. The relation between the 
probe qualities and the scanning operator can be expressed by means of Fourier's 
Integral Theorem so that either may be derived from the other. Examples of the ap- 
plication of the theory to record film and television scanning are discussed. 

The theory of scanning has been considered by a number of authors. 
The approach has, in general, been with respect to specific applica- 
tions in which the process of scanning has not been greatly differen- 
tiated from the kind of thing scanned. 

Scanning in itself is fundamental to most objective investigation. 
Thus the determination of a physical quantity ordinarily involves 
the use of a scanner or probe of some kind in which the effect upon the 
probe is taken as a measure of the quantity under investigation. 

It is evident that in general the process of scanning or measuring 
will give a result that is a kind of average of the region occupied by 
the probe, thus giving rise to an error in the resultant measurement. 
The magnitude of this error depends, in part, upon the relative di- 
mensions of the probe and the region of which a measurement is de- 

In this paper an attempt will be made to present a fundamental 
theory from which the various scanning relations can be derived as a 
natural consequence. 


Suppose for the moment that the region under consideration is one- 
dimensional and that its state can be represented by the scalar func- 
tion/^). This field is to be thought of as being contained in a line, 
with a line segment as the probe or scanner. 

* Received January 1, 1941. 
** WCAU Broadcasting Co., Philadelphia, Pa. 


<$ The Society is not responsible for statements by authors $ 



[J. S. M. P. E. 

The resulting measurement is proportional to a combination of the 
segment field and the field that the probe covers. The manner in 
which these fields combine depends upon their nature. It will be 
assumed throughout that the resultant field at any point is propor- 
tional to the product of their values at that point. 

A representation of this kind of arrangement is shown in Fig. 1, in 
which the segment field is shown graphically as an extension at right 
angles to the segment in the plane of the paper. 

The effective field F(x) included by the line segment is 

[)/(* + a)da 


where P(a) is the segment field and the integration is to be taken over 
the length of the segment. 

FIG. 1. One-dimensional representation. 

Consider the expression 

/! = /(* + ) 

and expand it in a Taylor series in a, 



Writing p = , equation 3 becomes, symbolically, 

where e ap is a transfer operator that shifts /(#) a distance a along the 

x axs. 

May, 1941] 


Scanning Operator 


T( P ) 

= sinh (ap} 

T(p) = sinh (ap) 

FIG. 2. Table of scanning operators. 

500 S. SABAROFF [J. S. M. P. E. 

Putting f(x + a) in the operational form in equation 1 results in 

F(x) = I P(a)**da /(*) (5) 


where the bracketed expression in equation 5 is an operator acting on 


= \ 


Then equation 5 is 

F(x) = 7W() (7) 

The evaluation of T(p) from equation 6 is an ordinary integration 
with the operator p regarded as a constant. The expression T(p) 
could be termed the scanning operator or scanning function, since its 
form is determined solely by the qualities of the scanning segment. 
A short table of scanning functions for some segment distributions 1 - 2 
that have appeared in the literature is shown in Fig. 2. 

Ordinary operational methods 3 ' 4 hold for the solution of equation 7. 
An important application is when f(x) is periodic in x. If the fre- 
quency in x is pi t the substitution of 

P = jpi (A) 

in equation 7 will give the complete solution. Equation 8 is utilized 
when the field being scanned is given in the form of a Fourier series 
or a Fourier integral. 

Another useful aspect of the operational method is the property of 
inversion. If in equation 7, F(x) and T(p) are known, then f(x) is 

/(*) = T-i(p)F(x) (9) 

where now the operator T~ 1 (p) is taken as acting on F(x). Thus the 
influence of the segment in F(x) can be cancelled by an application to 
it of the inverse of the scanning operator. 

The use of Fourier's Integral Theorem enables the determination 
of the segment field distribution for a given form of T(jpi). 

Rewrite equation 6 so that 

May, 1941] 



where jpi has been substituted for p. The length of the segment has 
been taken as extending through plus and minus infinity since the 
field can be considered as. zero outside of the actual segment extent. 
Equation 10 is a Fourier transform. 5 Its mate is 

thus giving the required segment field distribution. 

Examples illustrating the use of one dimensional theory are con- 
sidered in detail in Appendix 1. 

FIG. 3. Two-dimensional representation. 

The discussion thus far has been concerned with a one-dimensional 
theory. The theory for two dimensions can be similarly set forth. 

Let the field under consideration be represented by the scalar field 
f unction /(#, y). This field is to be scanned or investigated by means 
of a probe in the shape of a spot having a certain area and a field 
function given by P(<x, /3). a and are measured from a point of 
reference on the spot having the coordinates x and y. 

The scanning spot is supposed to lie on the field being scanned. A 
representation of this kind of an arrangement is shown in Fig. 3. 

The effective field F(x, y) included by the spot area is 

P(x, y) = 

f (P(<*. 

/*)/(* + , y + 


By methods similar to those used in the derivation of equation 4, 
the operational form for f(x + a, y -f /3) can be shown to be 

502 S. SABAROFF [j. S. M. P. E. 

f(x + a, y + /3) = eP + 0Qf(x, y) (13) 

where q = . 

Equation 12 can then be written as 

F(x, y) = T(p, q)f(x, y) (14) 

where the scanning function is 

C C 

T(P> j) ~ I ?( &) eap + Pidadp (15) 

J J 

As in the one-dimensional case, the evaluation of T(p, q) from equa- 
tion 15 is an ordinary integration with the operators p and q regarded 
as constants. T(p, q) may then be applied to f(x, y) in accordance 
with the usual operational theory. 

Of special importance is the case when f(x, y) is periodic with a fre- 
quency in x of pi and in y of qi. The substitution of 

p = jpi \ (16) 

2 = m / 

in equation 14 will then give the complete solution. 

The use of Fourier's Integral Theorem for two variables enables the 
determination of the proper spot field distribution for a given form 
of T(jpi, jqi) . 

Rewrite equation 15 so that 

+oo /+ oo 





where jpi and jq\ have been substituted for p and q, and the limits ex- 
tended to plus and minus infinity. 

Equation 17 is a Fourier transform for two variables. Its mate is 

P(ot, ft) = - I T(jpi, jq\}e~J<*Pi ~ ft<& dp\dq\ (18) 

thus giving the required spot field distribution. 

Solutions for fields that vary in two dimensions are not usually 
simple. In most practical applications, however, the scanning spot 
has circular symmetry, thus eliminating spot orientation complexities. 
The expression for T(p, q) for this case is not difficult to derive. 

The element of area for circular symmetry can be taken as sdsdO 
where a = S cos (6) and = 5 sin (0). If the spot field is P(S) at a 
radial distance S, then equation 15 becomes 

May, 1941] SCANNING THEORY 503 

T(M) = f 4 C 2 SP(S)eSM cos (O-tfdSde (19) 

Jo Jo 

where M = Vp 2 + q 2 , cos (</>) = />/M and sin (0) = 
Integrating equation 19 with respect to 6 gives finally 

T(M) = 2ir SP(S)I Q (SM)dS (20) 


where I Q (SM) is a modified Bessel's function of the first kind and zero 

When jMi is substituted for M in equation 20, it becomes a Fourier 
Bessel Transform, 6 thus 

= 27T f, 


$P(S)Ji($Mi)dS (21) 


where J Q (SMi) is an ordinary Bessel's function of the first kind and 
zero order. 

The mate of equation 21 is 

i f* 
p (S) = 2^- I MiT(jMi)J Q (SMi)dMi (22) 

thus enabling the determination of the spot field distribution for a 
given form of T(JMi). 

An example illustrating the application of two-dimensional theory 
is briefly discussed in Appendix 2. 


As yet no need of a scanning theory for more than two dimensions 
has presented itself. It must be pointed out, however, that the above 
theory can be extended to include a space of any number of dimen- 

For example, in the case of three dimensions, the probe may be 
thought of as a speck moving about in the region under investigation. 
If the field of the speck alone is P(a, ft, y) and the field being scanned 
is f(x, y, z) , then the effective field enclosed by the scanning speck is 

F(x, y,z) = \ \ \P(a, 0, y)f(x + a, y + 0, 2 + y)dctd(3dv (23) 

Operationally equation 23 can be written 

F(x. y, z) = T(p, q, r)f(x, y, z) (24) 

504 S. SABAROFF [j. s. M. P. E. 

The scanning function is 

, Q,r) = \ \ \P(a, ft, 

T(p, Q,r) = P(a, ft, y^P+dQ+yrdadpdv (25) 


where r = , 7 is the probe coordinate in the direction of z, and the 

other symbols remain the same as before. 

Equations 23, 24, and 25 may be written concisely by changing to 
vector notation. Let x, y, and z be the components of the position 
vector R, and a, (3, and 7 the components of the position vector 6, then 
equation 23 becomes 

F(R) = fp(tf(* + e)dv (26) 


where dv is a differential volume and P(e) B,ndf(R -f e) are scalar field 

Operationally, 26 is 

F(R) = r(v)/(*) (27) 

The scanning function is 

r(v) = fp(e)-v<fo (28) 


where V is a vector differential operator of space 7 and c.V is the scalar 
product of e and V . 

Disregarding the practicability of the matter, a complete three- 
dimensional scanning theory can be built up by the use of equations 
23 to 28, inclusive. 


The theory just outlined has been presented from a purely spatial 
point of view in which the only variable factor was the position of the 
scanning probe with respect to the field being scanned. Residual ef- 
fects have been taken as being entirely absent; that is, the past 
history of a scanning operation has been assumed as not affecting 
either the present or the future scanned output. 

In certain applications, the field scanned and the position of the 
scanning probe are both functions of time. 8 Ordinarily the probe 
size is small and the velocity of the scanning probe is so great that the 

May, 1941] 



field being scanned does not vary appreciably during a scanning opera- 
tion. This allows the application of the theory as an approximation. 

In television, which bases its operation upon the integrating charac- 
teristics of the eye, the past history of a scanning operation, and the 
probe field and size are of extreme importance. These are qualities 
that affect the fine structure and definition of television images. 1 
The theory may easily be modified in its application when these mat- 
ters are considered. 

Another application of the scanning theory is that it forms the basis 
of a spatial filter theory that is somewhat analogous to electrical 
filter theory. An example of such an application is considered in 
some detail in Appendix 1. 


FIG. 4. Rectangular scanning beam. 
Appendix 1 


The reproduction of sound from a film record utilizes the relative motion of a 
rectangular beam of light and a film of variable area or density. The light pass- 
ing through the film generates an electric current in a photocell which is brought 
to a desired level by means of amplifiers. It is of paramount importance to know 
the relation between the transmissivity of the film, the dimensions of the light- 
beam and the amount of light transmitted. This problem, in which the trans- 
missivity has been assumed to vary in a periodic manner, has been treated at length 
in the past. 9 ' 10 It will here be considered in the light of the foregoing theory. 

In Fig. 4, the scanning beam of semiwidth a and height A' is represented by 
ABDE, the measure of film transmissivity is denoted by/(x), the amount of light 
transmitted is measured by the area ABFG, and the relative position of the beam 
and film is given by x. 

506 S. SABAROFF [J. S. M. P. E. 

The scanning function is 

T(p) = K \ePd a (29) 

J -a 

where P(a) has been taken as equal to a constant K. 
Performing the integration in equation 29 gives 

T(p) = K (e*P - e-aP) = ^ sinh (ap) (30) 

P P 

The form of f(x) has not yet been specified. If it can be represented by a 
Fourier's series with a fundamental frequency of pi, thus 

n = oo 

f(x) = A + V A M cos (np lX + 6 n ) (31) 

where A n and n are the usual Fourier coefficients, then the application of 
equations 30 to 31 gives the well known result, 

^^^sinJpM CQS 

where the frequency of the constant term is taken as zero. f(x) may be discon- 
tinuous, and representable by the Fourier integral 


cos [pix + <f>(pi)]dpi (33) 

In the interval dpi, equation 33 is periodic in x with a frequency p\. Application 
of equations 30 to 33 therefore gives 



F(x) = 2Ka (fc) l cos [p lX + 4>(pdWpi (34) 

Jo api 

a result that has been arrived at in another way by Horton and Mathes. 11 
Equation 30 may be expanded in powers of p: 

2Ka y; (55) 

which when applied tof(x) results in 

Equation 56 is a form especially useful when/(jc) is given in the form of a series 
in x. 

In a certain densitometric application, a continuous density range can be 
scanned by means of a rectangular light-beam in order to obtain a measure of the 
variation in relative transmissivity. Obviously, any non-linearity in the density 
variation will cause a distorted relation by virtue of the aperture effect. F(x) 
and T(p) are therefore known and it is required to find f(x). 

May, 1941] SCANNING THEORY 507 

Operationally the problem is 

/(*) = T->(p)F(x) (37) 

where, from equation 30, 

T~*(p) = - csch (ap) (38) 


When F(x) is analyzed in terms of a Fourier series such as equation 31, the solu- 
tion is immediately 

M= oe 

2 napiA n csc(napi) cos (npix + 0) | (59) 


n ) 

If F(x) is obtained as a power series in x, it is advantageous to expand equation 
38 in a power series in p and then apply it to F(x) term by term as in equation 36. 
When this is done there results 12 

where 5 n are the Bernoulli Numbers 1 / 6 , 1 /so, 1 /42, etc. 

When equation 40 does not terminate finitely, its convergence must be investi- 
gated. A solution for/(#) when F(x) is of the third degree has been determined, 
and is 



F(x) = Fo + Fix + F&* + F*x* (42) 

Another solution for/fa) further illustrating the use of operational methods can 
be obtained. A partial fraction expansion for csch(op) is 13 

By the use of equations 38 and 43, 37 becomes 

Application of a typical term of the indicated summation in 44 to F"(x') results 

where X is a variable of integration. 

508 S. SABAROFF [j. S. M. P. E. 

The complete solution is finally 

The choice between equations 40 and 46 is dependent upon the form of F(x). 


It is evident that in the above example, the scanned output does not bear a 
linear relation to the field scanned, except perhaps in certain simple fields. This 
is entirely due to the shape and field distribution of the scanning beam. As may 
be seen from equation 32, the linearity increases as the beam width is decreased, 
with the disadvantage of a decrease in the scanned output. It would certainly be 
of interest to investigate the possibilities of a beam field distribution that would 
introduce no distortion in the scanned output, at least up to some predetermined 
cut-off frequency. 

The requirements are therefore, 

5 ,,;; w > 

where <o is the cut-off frequency and z is constant. 

Inserting the conditions of equation 47 in equation 11 gives 

() = ~ f 


Equation 48 is easily integrated, resulting finally in 

P() = A sin(ow) (49) 


Evidently the required scanning beam extends to plus and minus infinity. The 
beam intensity decreases in an oscillatory manner through positive and negative 
values from a maximum at the origin to zero at infinity. Such a beam is in 
reality not obtainable because it is impossible to secure a beam of infinite extent. 
The fact that the scanning beam is actually finite in width is usually the greatest 
limiting factor in securing any required frequency characteristic. The means for 
simulating negative fields in the scanning beam will not be discussed here. 

Assume that the field of the beam is zero outside of =*= and that the field is de- 
scribed by equation 49 within =*=a. From equation 10 the form for T(jpi) is 

r +0 si 
J a 

Because of symmetry about the origin, equation 50 may be written 

T(jpj = * f a sin (*) cos (fr) da (51} 

IT JO a 

After expanding the integrand in equation 51 into a trigonometric sum, re- 
writing and integrating, there results 

May, 1941] 



*\Si(a(u + pi}] -f 5t[o( - Pi)]l (52) 

< i 

where Si is the term for integral sine. 14 

A plot of equation 52 has been made in Fig. 5 in which aco has been taken as 
equal to 2rr, and z equal to unity. It is interesting to note the evidence of a defi- 
nite low-pass characteristic, even for this short beam-width. 

In certain applications, such as in television, it does not seem possible to simu- 
late a scanning beam with a negative field. However, suppose another field 
Pi(tt) is added to equation 49 such that their sum is never negative, thus: 




P( a ) = _5_ sin (aco) + P^a) 














IG. 5. Frequency characteristic, sine distribution. 

For convenience, the distribution described by equation 49 could be termed a 
"sine distribution," and that of equation 53 a "partial sine distribution." 

The form of Pi(a) should be so chosen that the qualities of the sine distribution 
may be as closely approximated as desired. This requires the contribution of 
Pi(a) to the scanned output to be as small as possible, at least in the region con- 
taining the frequencies of interest. These requirements are not usually easy of 

As an example, the effect of adding the sine distribution to the rectangular beam 
will now be considered. Assume that the distribution within a = =*=a is 


where k is a constant such that P(a) is never negative. 

The largest negative value of the sine distribution occurs when aw = 3ir/2, 
thus the smallest value that k can have is 2 /(Sir}. For this example it will be as- 
sumed that aw = Sir/2, and that k = 2/(37r). The relative distribution is shown in 



[J. S. M. P. E 

Fig. 6(a). The distribution is symmetrical about the center of the beam. In the 
figure, however, only the right half is shown. 

When equation 54 is inserted in 10, integrated, and the assumed values inserted, 
there is obtained 



PC*)- fSLATl/B 
o P p 
*> > OP 






HE Disrrte 













0.5- at a? 0.8 o.* /-o 

FIG. 6. Partial sine distribution. 

The relative values for T(jpi) have been plotted in Fig. 6(6). Quite evidently 
the low-pass characteristic has been somewhat obliterated by the addition of the 
rectangular beam. The accuracy of the curves is governed by tabular and slide- 
rule error. 

In concluding this section, it must be pointed out that the examples were 
considered mainly as illustrations of the application of the theory. In a similar 
manner beam-field distributions can be devised that will yield any required fre- 
quency characteristic. 

May, 1941] SCANNING THEORY 511 

Appendix 2 


A commonly used scanning spot is that generated by a cathode gun. This spot 
is circularly symmetrical with a radial field given by 16 

PCS) = Ve~a*S* (56) 

where V and a are constants. 

Inserting equation 56 in 20 and integrating gives the scanning operator 

T(M) = eM*/*a* (57) 

a 2 

If, as is customary, the field being scanned is represented by the double Fourier 

/(* y) - SS^ nm COS \ U P& + m &y + 8 nm] (58) 

n m 

and equation 57 is applied, there is obtained 


Examination of equation 59 shows that the distortion due to scanning in this 
case is caused by a continuing diminution in amplitude of the higher order har- 


As in Appendix 1, an examination of the possibilities of a distortionless scanning 
spot, at least up to some predetermined cut-off frequency, would certainly be of 

The requirements are 

! z ' Ml < " 
\0, M, >o> 

where o> is the cut-off frequency and z is constant. 

It must be remembered that in terms of the Fourier field components of (58), 

M, = (fc + wV) 1 /. (61) 

The wavelength of a field component is 2ir/M } ; thus the conditions of equation 
60 restrict the spot response to components with wavelengths greater than 2ir/w. 
Putting equation 60 into 21 gives 

P(S) = A I M l J (SM l )dM l (62) 

Equation 62 is easily integrated, resulting in d 


512 S. SABAROFF tf. s. M. P. E. 

The required spot field extends to infinity with an intensity that decreases in an 
oscillatory manner through positive and negative values from a maximum at the 
origin to zero at infinity. 

Such a condition is not exactly obtainable because in this application it does not 
seem possible to simulate negative spot fields. Also the spot is in reality of finite 

It is possible to add another field Pi(S) to equation 63 such that their sum is 
never negative: 

P(S) = /!(5) + Pi(S) (64) 

Equation 63 could be termed a "Bessel distribution" and equation 64 a "partial 
Bessel distribution." 

The frequency characteristic of the partial Bessel distribution may be found by 
inserting equation 64 in 21, i. e., 

= zu C a j l (S^)J Q (SM l )dS + 2x C 
Jo Jo 

SPi(S)J (SM l )dS (65) 

where a has been taken as the radial extent of the spot. 

It is interesting to note that the first integral in equation 65 is a step function 16 
such that 

No tables of equation 66 for a finite upper limit seem to be available. Rough 
calculations show it to be somewhat similar to the frequency characteristic of the 
sine distribution discussed in Appendix 1. 


1 WHEELER, H. A., AND LOUGHRAN, A. V. : "The Fine Structure of Television 
Images," Proc. I. R. E., 26 (May, 1938), pp. 540-575. 

2 GOLDMARK, P. C., AND DYER, J. N. : "Quality in Television Pictures," 
Proc. I. R. E., 28 (Aug., 1940), pp. 343-350. 

3 STEPHENS, E.: "The Elementary Theory of Operational Mathematics," 
McGraw-Hill Book Co., Inc. (1937). 

4 DAVIS, H. T. : "The Theory of Linear Operators," Principia Press (1936). 

5 MACROBERT, T. A.: "Functions of a Complex Variable," Macmillan Co. 
(1933), pp. 319-321. 

*Ibid., p. 321. 

7 COFFIN, J. G. : "Vector Analysis," John Wiley & Sons, Inc. (Second Edi- 
tion, 1911), p. 131. 

8 MERTZ, P., AND GRAY, F.: "A Theory of Scanning and Its Relation to the 
Characteristics of the Transmission Signal in Telephotography and Television," 
Bell Telephone System, Mon. B-799. 

9 STRYKER, N. R.: "Scanning Losses in Reproduction," /. Soc. Mot. Pict. 
Eng., XV (Nov., 1930), pp. 610-623. 

May, 1941] SCANNING THEORY 513 

10 COOK, E. D.: "The Aperture Effect," /. Soc. Mot. Pict. Eng., XIV (June, 
1930), pp. 650-6f>2. 

11 GRAY, F., HORTON, J. W., AND MATHES, R. C. : "The Production and Utili- 
zation of Television Signals," Bell Syst. Tech. J., 6 (Oct., 1927), pp. 560-603. 

12 DWIGHT, H. B.: "Tables of Integrals," No. 657.8, Macmillan Company 
(1934), p. 132. 

13 Ibid. (5), ex. 2, p. 104. 

14 JAHNKE, E., AND EMDE, F.: "Tables of Functions," B. G. Teubner, Leipzig 
and Berlin (1933), pp. 78-86. 

16 ZWORKIN, V. K., AND MORTON, G. A. : "Television," eq. 13.8, John Wiley & 
Sons, Inc. (1940), p. 372. 

16 Ibid. (5), ex. p. 323 due to Sonine. 


Summary. For the most part, since the advent of talking pictures, motors have 
been employed whose performance is excellent. The various types of motor, however, 
differ widely in their ability to resist load irregularities and in their tendency to 
oscillate when a disturbance occurs. For the more critical applications these fac- 
tors deserve careful consideration when the type or design is being selected. The prin- 
cipal types of synchronous motor are (1} the variable-reluctance or induced-pole 
motor, (2} the separately excited motor, (5) the a-c-d-c motor, (4) the hysteresis motor, 
(5) the low-speed multi-tooth motor (of the type used for electric clocks}, (6) the poly- 
phase, uniform-torque modification of number 5, and (7) selsyn motors. 

Many of the characteristics of synchronous motors may best be understood by as- 
suming that the polyphase winding produces a uniformly rotating magnetic field, 
but estimating the stiffness and stability demands a knowledge of the manner in which 
the a-c input varies with mechanical displacement. Generous pole-face grids are 
essential for stability. A-c-d-c motors have certain elements of instability as well as 
stabilizing factors, which are not present in straight synchronous motors. The mag- 
nitude of these effects can to some extent be controlled by the external circuit arrange- 
ments. Selsyn motors are less readily damped than regular synchronous motors, 
and for this reason arrangements by which the synchronous motors can be interlocked 
from standstill are of interest. 

In the days of the silent motion picture, there was nothing critical 
about the speed at which the film was run. In fact, prevailing pro- 
jector speeds had gradually crawled upward from 60 to 70 or 80 feet 
per minute and most any kind of motor that would not stall was good 

With the advent of sound, the adoption of a standard and rigid 
adherence to the same was immediately essential. Engineers promptly 
realized this and after a short period of question as to what the 
standard speed was to be, adopted 90 feet per minute as a standard 
and fell to work to build accurately controlled driving systems. 
These included governed d-c motors and motors controlled by electric 
governors which were highly sensitive to the frequency of a tone 

* Presented at the 1940 Fall Meeting at Hollywood, Calif. ; received February 
28, 1941. 

** RCA Manufacturing Co., Caniden, N. J. 


*>The Society is not responsible for statements by author s& 


generated by a device attached to the motor. Where a power source 
of reasonably constant frequency was available, synchronous motors 
were used and these have been highly satisfactory. Induction motors 
have been widely used for machines which did not have to be syn- 
chronized with others, and where a variation of one or two per cent 
was not serious, provided there were no rapid changes. The average 
slip of an induction motor can be compensated in the gearing. 

Synchronous motors for use in the motion picture industry are for 
the most part designed to work with polyphase alternating currents. 
In an appropriately designed winding, the polyphase currents result 
in a magnetic field of practically constant intensity, which rotates 


FIG. 1. Production of a rotating magnetic field by 
a polyphase winding. 

with respect to the windings at a speed determined by the number of 
poles which the winding is designed to produce, and the frequency of 
the a-c supply. Fig. 1 shows the current distribution in a simple two- 
phase winding, and the resulting magnetic field at several stages of a 
cycle of alternation. The progressive shift of the field (which means 
continuous rotation) is clearly evident in the figure. The rotor is so 
designed that it follows this moving magnetic field, as if flexibly tied 
to it, but without any continuous drift behind, or slip. There are 
several types of synchronous motor which will be briefly described. 
Fig. 2 shows one of the earliest types of synchronous motor. Here 
the synchronous speed obviously depends on the local power plant. 

The Variable-Reluctance or Induction- Synchronous Motor. The 
type of synchronous motor which has been most widely used from the 
start is practically a polyphase squirrel cage induction motor, in the 


S. READ, JR., AND E. W. KELLOGG [J. S. M. p. E. 

rotor of which some of the teeth have been cut away or shortened to 
give the effect of salient poles. Fig. 3 shows the approximate rotor 
cross-section of such a motor. The motor is usually started by apply- 

A -Full Load B- No Load C-Rctarct. 

FIG. 2. Early type of synchronous motor: The 
reacting gravitational field rotates backward with re- 
spect to the rotor at synchronous speed but is stationary 
in space. The three views illustrate conditions at 
different stages of oscillatory cycle. 

ing full voltage to the polyphase primary. It quickly comes up to 
nearly synchronous speed, acting entirely as an induction motor. 
When the speed is close enough to synchronism, the salient poles lock 
in with the rotating magnetic field produced by the polyphase wind- 
ing and thereafter the motor 
continues to run in strict syn- 
chronism. Because the torque 
developed in a motor of this kind 
depends on having low magnetic 
reluctance in line with the field- 
poles and high reluctance be- 
tween the poles, it is commonly 
called a "variable-reluctance" 

Inter changeability of Rotor and 
Stator. In most designs of syn- 
chronous motors, the a-c winding 
may be on the stator and the field 
may be the rotor, or the field may 
be stationary and the alternating 
currents supplied to the rotor through slip-rings. The principles of 
operation are in no wise altered by an interchange, but practical de- 
sign considerations will favor or perhaps necessitate the choice of one 
or the other. In general, it is desirable to have the field rotate since 

FIG. 3. Manner in which rotor 
teeth are cut away to produce salient- 
pole effect. 

May, 1941] 



this calls for only two slip-rings at most, and in some constructions 
none at all. The stator or outside member, of course, affords more 
room for the windings. 

A-C^D-C Motors. The voltage between any two commutator 
bars of a d-c motor is alternating. If slip-rings are connected to 
suitably located taps in the armature winding, the machine becomes 
a rotary converter or an a-c-d-c motor. Fig. 4 shows schematically 
the armature of a two-pole rotary converter. The ring type of arma- 
ture is used for illustration because the relation of the active (outside) 
conductors to the field, brushes, and slip-ring taps is most easily seen 
in the diagram of such a winding. The difference between the ring 
winding and the usual practical winding is that in the latter the con- 
ductors, instead of returning through a hole in the armature, cross it 
diametrically and return through 
the bottom of a slot under the op- 
posite pole, thus doubling the 
number of effective conductors. 

In a-c-d-c motors, the armature 
is necessarily the rotating member, 
with the field stationary. As ap- 
plied to sound-recording work, a-c 
-d-c motors may be employed as 
straight a-c synchronous motors, 

FIG. 4. Elementary type of a-c- 
d-c motor or rotary converter. 

or the power may be supplied from 
the d-c side and the a-c connec- 
tions employed solely for the purpose of keeping several machines 
running at identical speed. Some of the well recognized character- 
istics of rotary converters apply to the a-c-d-c motor, but in inter- 
preting any textbook statements about rotary converters, it should 
be borne in mind that in general the latter carry no mechanical load, 
the power intake on one side being expended almost entirely to put 
out electrical power on the other side. Under these conditions, the 
distortion of the field by the armature currents is much less than in 
the case where power intake may be from either or both sides and ex- 
pended in overcoming mechanical loads. 

Motors with D-C Field. Large synchronous motors and generators 
always have field windings and d-c excitation. This, of course, 
means complication, which is particularly undesirable in small ma- 
chines, the induced field or variable reluctance motor being ideal from 
the standpoint of simplicity. The presence of a separately excited 

518 S. READ, JR., AND E. W. KELLOGG IJ. S. M. P. E. 

field reduces the a-c input requirements, largely eliminating the low 
power-factor demands of the motor. Other factors being equal, the 
addition of a field- winding around the salient poles of a field tends to 
stiffen the connection or make the motor drop back less in phase 
when a mechanical load is applied. 

Permanent-Magnet Fields. It might appear that since the advent 
of such potent materials for permanent magnets as "alnico" or similar 
alloys, it would be a simple matter to provide a rotating field with 
the required excitation by employing such magnetic material, and 
thereby to obtain a motor which combines the advantage of stiffness 
and high power-factor with the simplicity of the variable-reluctance 
motor which uses a soft steel rotor core. There undoubtedly are 
possibilities in this direction, but in ordinary service, small motors 
often have to stand full applied voltage when out of synchronism, 
as, for example, during starting, and at such times the field would be 
subjected to very strong demagnetizing action. This would alter- 
nate with moments of magnetic reinforcement from the a-c winding, 
but the result would be that the field would always be somewhat 
weaker than the desired normal working value, and the difference 
would have to be made up by lagging primary currents. These 
would have to make up the difference in field strength, but have to do 
so through a field structure comprising material which is magnetically 
very hard. The reduction of the required excitation which the a-c 
winding must supply would thus be less than might at first be ex- 
pected. The foregoing applies to motors with rotating field. Per- 
manent magnets, on the other hand, have frequently been applied to 
stationary fields. Here a large leakage factor can be allowed and the 
demagnetizing effects of the a-c winding have less effect on the per- 
manent magnet material. 

Hysteresis Motor. Although there appears to be little advantage 
in using permanent-magnet material in a motor having a salient-pole 
rotor, there is a type of synchronous motor which has found many 
applications which employs magnetically hard material in the rotor. 
This is sometimes called a "hysteresis motor." The polyphase stator 
produces a rotating magnetic field. The rotor is made of circular 
punchings of magnetically hard material, but depends on being mag- 
netized by the stator currents rather than on a previous magnetiza- 
tion. If the rotor is held still while polyphase alternating current is 
applied to the stator, the rotor is continuously remagnetized in chang- 
ing direction. The reaction of its residual magnetism with the mag- 


netic field from the stator exerts a forward torque which is practically 
fixed in magnitude and independent of speed or slip. Such a motor 
gives constant torque up to actual synchronism. In other words, it 
can develop its full possible torque without any slip whatever, and if 
the load calls for less than the maximum torque which the motor can 
supply it will operate in strict synchronism. The constant-torque 
characteristic is based on the assumption that the motor is not pro- 
vided with any secondary or squirrel-cage winding. If a secondary 
winding is provided, it will result in additional torque at all speeds be- 
low synchronism. The machine then acts as an ordinary induction 
motor but with the hysteresis torque added to that due to the in- 
duced currents in the secondary. In the absence of any squirrel-cage 
winding, the hysteresis motor is not very well damped. There is 
large energy loss if the oscillations are of sufficient amplitude to slip 
the magnetism in the rotor core but when the amplitude falls below 
a certain point small oscillations can persist without much damping. 

In the case of motors of the salient-pole or wound-field type, there 
are certain positions at which the motor must lock in, and this in some 
applications may be of importance for maintaining accurate phase 
relations between several machines. If such machines, due to any 
temporary excess load, should break out of synchronism or slip back 
in phase at all, they will usually stop entirely or take such excessive 
current as to give an indication of something being wrong. The hys- 
teresis motor, on the other hand, if momentarily loaded, can slip 
back in phase and no one be the wiser. For this reason (in part at 
least) it has not so far found much place in synchronous sound re- 
cording even though it is capable of carrying its normal load in per- 
fect synchronism. 

Low- Speed Toothed-Wheel Single-Phase Motors. In any list of 
types of small synchronous motors we must include the type which 
has found such wide application in driving clocks. These motors 
consist simply in toothed wheels, outside of which are similarly 
toothed rings or punchings, and a winding so arranged as to produce 
a magnetic field between the rotor and stator. The motor runs at 
such speed and in such phase relation to the current, that the inner 
and outer teeth are opposite at the time of maximum magnetizing 
force or winding current. If no polarizing is provided, the rotor 
moves one tooth pitch in Vi2o second. These machines must be 
started by an impulse, which may be supplied by hand or by some 
auxiliary device. 


S. READ, JR., AND E. W. KELLOGG [J. S. M. P. E. 

Self -Starting Toothed-Wheel Motors. By employing a polyphase 
winding or producing equivalent effect by "shading coils" the toothed- 
wheel, low-speed motor can be made self-starting and can be given a 
practically constant torque characteristic instead of jerks at V2o-sec- 
ond intervals. Fig. 5 shows such a motor, used in small phonographs. 
Selsyn Motors. A very important type of synchronous motor is 
the "selsyn." It is essentially an induction motor with wound sec- 
ondary instead of squirrel-cage construction, and requires slip-ring 
connections to the rotor. It is necessary that either the primary or 
secondary shall be polyphase, and for best operation both are poly- 
phase. In operation, polyphase alternating current is applied to 

either the rotor or stator windings 
of a number of machines, the other 
windings of the motors being in 
parallel but not connected to any 
source of power. So far as prin- 
ciple of operation goes, it makes 
no difference whether the power 
is applied to the rotor or to the 
stator windings, but in discussing 
the machines it will be conve- 
nient to assume that the pri- 
mary windings are the stators. 
When the voltage is applied to 
~ the stators (the rotors being 
connected together) the latter 
will undergo slight rotations and will seek such relative positions 
that very little current circulates. The machines in this condition 
act like a group of transformers paralleled on both the primary 
and secondary sides, but with no load on the secondary side. If the 
rotor of any machine is moved in either direction, the balance is up- 
set, and cross-currents flow in such direction that they pull back on 
the advanced rotor and pull forward on the other rotors. If the 
rotors are free to turn, the entire group act as if geared together. 
Driving any one of the machines will cause all the others to run at ex- 
actly the same speed. Obviously, if one machine is to control a 
number of others, it should be larger than the ones which it controls. 
In sound-picture work, selsyn motors are used where it is important 
to lock a number of machines together before any are started, and to 
start and stop the entire group without any changes in relative posi- 

FIG. 5. 

Polyphase low-speed 
chronous motor. 


tion, or, in other words, to make all the machines concerned feed the 
identical footage of film, including that which passes during the 
periods of acceleration and stopping. This is especially important in 
re-recording work where a number of reproducing machines are em- 
ployed to supply sounds which are to be mixed together. 

Elementary Theory of the Synchronous Motor. Many of the char- 
acteristics of synchronous motors can be interpreted and understood 
best by making the simple assumption that the polyphase winding to 
which alternating current is supplied produces, relative to the wind- 
ing, a magnetic field which rotates in space. This was illustrated in 
Fig. 1 . By suitable distribution of the windings in the slots, the con- 
dition of uniform field rotating with uniform velocity can be closely 
approximated. If a suitably shaped piece of iron is placed within 
the rotor, it will obviously attempt to follow the magnetic field 
around. The iron armature becomes itself a magnet and its reaction 
with an external field is such that it tries to line up with the field. If 
the rotor is otherwise magnetized, the same thing occurs, but the 
rotor now becomes more particular about which of its poles shall be 
under which magnetic pole of the stator. The speed at which the 
stator field rotates is determined entirely by the winding and the 
frequency of the supply. If the winding is such as to produce two 
north and two south poles within the stator and the supply is 60 
cycles, the field will rotate 1800 rpm. If the voltage is applied to 
the stator while the rotor is stationary, or if the rotor is running at 
some different speed, the rotor will at each instant try to line up with 
whatever stator pole happens to be nearest it, and it vacillates be- 
tween trying to follow the stator pole which is just leaving, and going 
back to meet the one which it sees coming. The former tends to 
pull it forward and the latter to pull it backward. The forward and 
backward pulls alternate with the passage of the magnetic poles, but 
are equal on the average and no net torque is developed. The syn- 
chronous motor so far described must, therefore, have some external 
means for starting, or be constructed so that it will have a forward 
torque at speeds below synchronism. The regular induction motor 
has such a forward torque and its features can readily be combined 
with those of the synchronous motor. Induction motors ordinarily 
have uniform air-gaps. A set of conductors is arranged in slots 
around the rotor, usually in the form of copper or aluminum bars con- 
nected to low-resistance rings at each end. This is known as the 
"squirrel-cage" construction. When the magnetic field from the 

522 S. READ, JR., AND E. W. KELLOGG [J. S. M. P. E. 

stator travels around the rotor, it cuts the conductors and induces 
currents in them (provided there is difference in speed) and these cur- 
rents react with the field in a direction to resist relative motion, or, in 
other words, to carry the rotor around with the field. The torque so 
developed depends on the speed difference or "slip." When the slip 
is large, the induced currents are large, but alternate at a frequency 
comparable with that of the supply. Leakage inductance causes 
them to lag, with the result that they do not exert as much torque as 
smaller currents which are in phase with the induced voltage. As 
the motor speeds up and the slip becomes less, the induced voltage in 
the rotor bars falls, but the phase difference becomes less, with the re- 
sult that the torque rises to a maximum. With still further reduction 
in the slip the fall in voltage and current is not compensated by 
higher power-factor, and the torque falls off. Although the induced 
voltage is low at small slip, the resulting current can still be fairly 
large if the resistance of the conductors is low, and there will still be 
good torque developed. For this reason, if the induction-motor ac- 
tion is required to bring a machine as nearly as possible to synchro- 
nism, the resistance of the rotor conductors must be kept very low. 

Nearly everyone has watched the oscillations of a compass when 
the needle is first set free to line up with the earth's field. The force 
with which a magnet seeks to line up with the magnetic field is like 
that of an elastic connection. In the case of a synchronous motor, 
the rotor poles will line up very closely with the regions of maximum 
magnetic strength due to the winding currents. As load is applied 
to the shaft, the rotor poles drop back in position but continue to 
follow at full speed, although slightly behind in phase; just as the 
draw-bar through which a train is pulled, pulls out (against spring 
action) when the train starts up hill, but the cars still run at the same 
speed as the locomotive. If some action causes the rotor poles to as- 
sume a position ahead of the stator magnetic field, they will be pulled 
back toward equilibrium position. The fact that the driving force 
is like an elastic spring and the armature has considerable mass, 
means that the synchronous motor is capable of oscillation or "hunt- 
ing." Fortunately, there are very effective means for damping 
which are applicable to most synchronous motors. The same 
squirrel-cage or rotor conductors provided to bring the motor up to 
speed serve also to damp oscillations. If the stator has been lagging 
behind its normal or average position, and thereby receives an excess 
of forward pull, the machine is accelerated and for a moment runs 


above normal speed to catch up. It then has an excess of kinetic en- 
ergy and over-runs until the stator is ahead of its normal position. 
The magnetic action then pulls it back, and it again overshoots to 
the lagging position. To damp the oscillation, a retarding force must 
be applied not at the moment of maximum forward deflection, but 
during the moment of overspeeding that precedes the maximum de- 
flection, and a forward torque is needed during the part of the cycle 
when the rotor is dropping back in phase, rather than at the time 
when it is farthest behind. The induced currents in the rotor con- 
ductors resist the shifting of relative positions of the rotor poles and 
the magnetic field with which they are reacting. Thus if the mag- 
netic flux tends to be dragged back and forth across the rotor poles, 
the induced currents in the bars of the grids resist these changes. 
Very powerful damping can be obtained by this expedient provided 
the conductors are of low enough resistance. 

Stiffness. The fact that the poles of the rotor and stator try to 
line up means that there is an elastic coupling, as already described. 
This would be true even though the armature currents were not af- 
fected by the applied load or the displacement from the in-line posi- 
tion. In practically all synchronous motors, the input current is 
strongly affected by a displacement in the relative armature and 
field positions, and the increased current if the deflection is increased 
is in such a direction as to resist the deflection. In other words, the 
fact that the current changes with deflection results in a great increase 
in stiffness. In order to see why and how this happens, we shall re- 
sort to vector diagrams. The method of analysis is one widely ap- 
plied to large machines with d-c excitation. Most of the conclusions, 
however, are at least qualitatively applicable to small machines, in- 
cluding those of the induced-pole type. 

Calculation of Currents. In Fig. 6 (a) the vector marked E t repre- 
sents the voltage impressed across one phase of the rotor winding. 
The vector E g represents the voltage induced by the conductors cut- 
ting the magnetic field. E g in general opposes the impressed voltage 
E t , the current flow being regulated by the difference between these 
two voltages. Hence the vectors E t and E g are drawn in approxi- 
mately opposite directions, but they are not exactly opposite. The 
effect of applying a mechanical load is to make any given armature 
coil arrive at the middle of a field-pole slightly later than it would 
without the load, and therefore to make the generated voltage E g 
fall back in phase. The angle of lag of E g is designated as x in Fig. 6. 


S. READ, JR., AND E. W. KELLOGG [J. S. M. P. E. 

If we are given the two voltages E t and R g and the angle x, we can 
easily find their resultant, which is shown as E s in Fig. 6. The resis- 
tance R and inductive reactance coL of the windings give them a certain 
impedance which we designate as Z, and the current / lags behind 
the voltage E s , which drives it through the windings, by an angle p 
which depends on the ratio of inductance to resistance. Having 
found E s we can therefore draw the current vector 7. 

The next question is how much torque does the current / produce. 
At any given motor speed the torque is proportional to the electrical 
power converted into mechanical power. This power is proportional 
to E g l cos y, in which y is the angle between 7 and E g . In Fig. 6 (a) 

FIG. 6. Voltage and current relations in synchronous motor. 

I cos y appears as the projection OD of vector I on the extension of 
the E g vector. 

Although the action of the machine may be analyzed by going 
through the steps just described, the diagram can be simplified and 
the effects of changing certain quantities more easily followed. In 
Fig. 6(a) a parallelogram is used to find the resultant E s of E g and 
E t . Only the lower half of the parallelogram is really needed. 
Thus in Fig. 6(6) E g is laid off from O as before, the line GS is drawn to 
represent the vector E t and the line OS which closes the triangle is 
E s . From E s , I and I cos y are found as before. 

If the immediate purpose of the diagrams is to study the effect of 
displacement x on input power, there is a further short-cut. In Fig. 
6(c) a reference axis MN has been drawn, making the angle p with 
the axis OD. The angle y now appears between E s and the new axis 

May, 1941] 



MN, and the projection of E s on MN is E s cos y. For given values 
of E g and Z, E s cos y is proportional to torque. Therefore to study 
the effects of changes in displacement angle x on torque, we need 
only note the effect of changing x on E s cos y, found by projecting E s 
on axis MN. Fig. 6(d) shows the quadrants in which E s may fall 
and whether these represent motor action or generator action and 
leading or lagging armature current. The leading or lagging com- 
ponents of current supply no power, but act either to oppose or to 
reinforce the main field excitation. 

With the foregoing theory as a background, let us consider the 
effect of displacement on power intake, first for the case of a motor 
having a high value of &L/R. Appropriate diagrams are shown in 

FIG. 7. Current vs. load relations in machine with high 
ratio of X/R. 

Fig. 7, in which diagram a represents the conditions when E g is less 
than E t , and diagram b the conditions when E g is greater than E t . 
In each case three values of displacement angle x are taken, X being 
the angle which causes no power to be supplied to the machine, X m 
is an angle giving motor action, and X g an angle which causes gen- 
erator action. Only one set of the vectors E g , E t , E s , and E s cos y 
in Fig. 7 (a) is labeled: namely, the one corresponding to X m . The 
locus of E t as x changes is an arc of a circle, and since the E 5 vector 
is drawn from the end O of the E g vector to the end T of the E t vector, 
its locus is the same circle. It will be noticed that the effect of chang- 
ing x is to cause a major change in E s cos y, for the reason that the 
circular arc, which is the locus of the end of the E s vector, is nearly 
parallel to the axis MN. On the other hand, the component of E s 
perpendicular to MN (which means lagging current in the case of 

526 S. READ, JR., AND E. W. KELLOGG [j. s. M. p. E. 

Fig. 7 (a), or leading current in 7b) changes only slightly with changes 
in load (changes in the angle x). The rapid increase in the torque- 
producing component of current (proportional to E s cos y) with change 
of displacement x, means that such a motor will strongly resist a dis- 
placement, or, in other words, it will be stiff. 

Effect of Resistance on Stiffness. Let us next, for the sake of com- 
parison, assume that resistance predominates over reactance in the 
motor-winding impedance, making the angle p small, which gives the 
reference axis MN a direction more nearly at right-angles to the cir- 
cular arc. Vector diagrams for a somewhat exaggerated case of such 
a machine are illustarted in Fig. 8. We now find that a shift in the 

FIG. 8. Current vs. load relations in machine with low 
ratio of X/R. 

displacement angle x between E g and E t results principally in produc- 
ing leading or lagging currents with comparatively small changes in 
the torque produced. Therefore, high resistance in the winding or 
the power-supply leads greatly reduces the stiffness of the motor. 
As compared with Fig. 7, the conditions of Fig. 8 (small value of 
uL/R) result in a machine in which mechanical load produces large 
differences in the leading or lagging components of current, and there- 
fore the field-strength is strongly affected by the load. It further 
appears from Fig. 8 that unless the generated voltage E g is considera- 
bly less than the line voltage E h the machine will not be able to carry 
much motor lead. If x m increases, the projection OF, which repre- 
sents power intake, will increase only slightly and then begin to de- 
crease as T m passes the tangent point. This means that the motor 
would fall out of step. If the generated voltage were 20 per cent 


higher than shown, the machine would not be able to run at all as a 
motor. It should be noted also that as the load is increased (ap- 
proaching the pull-out limit) the stiffness rapidly becomes poorer, 
and fluctuations in the leading (demagnetizing) component of arma- 
ture current with changes of load (i. e., changes in the length of FT m ) 
are exaggerated. 

In the analysis just given, E g is based on normal field-strength ex- 
cluding the effect of armature reaction on the field, but armature re- 
action is taken into account in the value taken for winding inductance 
(see Appendix). The method is not rigorously applicable to salient- 

FIG. 9. Testing motors for stiffness and oscillation by 
means of stroboscope. 

pole motors, because the armature reaction and therefore the value 
of Z is not the same for components of current which are in phase and 
those which are in quadrature with the voltage. The quadrature 
currents reach maximum when surrounding the main field, and in this 
position, the reluctance is low and a large magnetic effect is produced. 
The in-phase or power-carrying currents reach maximum when the 
coils are turned at right-angles to the field and the reluctance is maxi- 
mum. Therefore, the reactance is lower for the power components 
of armature currents than for the magnetizing components, and this 
reduces, though it does not eliminate, the detrimental effects of re- 
sistance on the phase-angle of the current. An analysis which takes 
the difference in reluctance into account is given in an Appendix. 


S. READ, JR., AND E. W. KELLOGG [J. S. M. p. E 

FIG. 10 (a). A-c-d-c motor, separate excitation // 
0.385; no resistance in leads. 

FIG. 10 (&). A-c-d-c motor, separate excitation // 
0.390; R = 3 3 /4 ohms in each lead. 

FIG. 10 (c). A-c-d-c motor, self -excited, // = 0.390 
ampere; R = 3 3 /4 ohms. 

The reduced stiffness resulting from inserting resistance in the leads 
may be observed by mounting a stroboscopic disk on the motor shaft, 
illuminating the disk with a flashing light-source and noting how 
far the armature is deflected when a small brake load is applied. 
Fig. 9 shows this test being made. A still more satisfactory method of 
studying the performance of a motor is by means of oscillograms of 
armature displacement, such as those shown in Figs. 10 and 11. 
The reduction in stiffness due to the resistances is shown by the lower 
frequency of the oscillations in curve b of Fig. 10 than in curve a. 
The tests were made on an a-c-d-c motor. The armature was de- 
flected 5 mechanical degrees and then released. (The 5-degree deflec- 
tion was reached with a much less brake torque load in some cases 
than in others.) When the resistances were introduced, the supply 

May, 1941] 



voltage was raised to make up for the drop in the resistors. The 
method of making the oscillograms is of interest. The output of a 
magnetic tone-wheel on the shaft of the motor being tested was com- 
bined with that of a similar reference tone- wheel on another motor in 
such a way that rectification gave a direct current bearing a linear 
relation to the angle of phase displacement between the motors, and 
this direct current was supplied to the oscillograph galvanometer. 
Each oscillogram covers a period of about 4 seconds. Measurements 
of stiffness were also made by means of a prony brake and a strobo- 
scope. The results are given in Table I. 


to D-C 


at Slip 



in Oz-In per 
No At 75 
Load Oz-In 



















3 3 / 4 ohm 








3 3 /4 ohm 
















3 3 / 4 ohm 








3 3 / 4 ohm 














* D-c line voltage equal to motor brush voltage before connecting. 

Effect of Field Excitation on Stiffness. It is of interest to compare 
the stiffness of a motor provided with d-c excitation with a salient- 
pole motor depending entirely on the a-c winding to supply the excita- 
tion. Obviously, the stiffness of the latter depends on how much the 
reluctance is increased between poles, or, in other words, on how much 
of the field structure is cut away. With a d-c field winding, on the 
other hand, a substantial stiffness can be had with a uniform air-gap. 
The only condition, therefore, in which the two types of motor would 
be comparable is that in which the reluctance between poles is so high 
and the flux in this region so weak that it exerts little force. If the 
resistance and leakage reactance of the armature windings are small, 
the total field flux linking the armature turns will be nearly the same 
in two motors of similar proportions and a-c winding design, one of 
which has d-c excitation, while the other depends on the a-c winding 
to produce the field. Those who have wrestled with transformer 
theory will recall that the general design formulas assume a definite 


S. READ, JR., AND E. W. KELLOGG [J. S. M. P. E. 


FIG. 11 (a). A-c-d-c motor, field'winding shorted. 

FIG. 11 (6). A-c-d-c motor, self -excited, // = 0.29 
ampere, 1800 rpm. 

FIG 11 (c). 

A-c-d-c motor, self-excited, // 
ampere. * 


relation between the applied voltage, the number of turns, and the 
maximum magnetic flux. A similar relation holds in the case of the 
synchronous motor. The total amount of flux which a given wind- 
ing must interlink in order to develop a counter-elctromotive force 
substantially equal to the impressed voltage is determined by the 
number of turns in the winding. The general conclusion is that if a 
separate winding does not supply the necessary field, then the a-c 
winding will do so. It accomplishes this, of course, by drawing a 
strong lagging current, and while this, in certain applications, may be 
considered a drawback, it does not seriously prejudice the operation 
of the motor, provided the winding is of low resistance. 

The principal difference between the stiffness of a motor with field 
excitation, and one with similar field structure, but depending on the 
a-c winding to provide excitation, is that the force reaction between 
the magnetizing component of current and the cross-field is opposite 
in sign to the reaction between the power component of current and 
the main field. Therefore eliminating the magnetizing current re- 


moves this subtractive factor. Another way of describing the same 
effect is to say that with d-c excitation the resultant field does not 
shift in position with changing load as much as it does in the case of 
the induced-pole motor. If a given motor is tested (as illustrated in 
Table I) with and without field excitation, there is further loss of 
stiffness for the case of a-c excitation, due to the a-c winding impe- 
dance and consequent somewhat reduced field-strength. On the other 
hand, since iron losses and saturation are the factors which set the 
flux limits, it is logical to design a simple variable-reluctance motor to 
operate with about the same flux density as a d-c excited motor of 
like proportions. Thus in comparing types of motors, the field- 
strength factor should not be counted, but in the case of small motors 
a difference in stiffness may be expected of the order of 1.5 to 1 in 
favor of the motor with d-c field, due to the fact that the field does 
not shift as much with changes of load. How much this difference 
amounts to depends largely on the ratio of reluctance in line with and 
at right-angles to the main field-poles. 

Fig. 1 1 shows the effect of suddenly removing a brake load from the 
a-c-d-c or "interlock" motor with which the tests of Fig. 10 were 
made. It is seen that the natural period (which is determined by the 
stiffness) is only slightly altered by excitation. On the other hand, 
the tendency to oscillate increases decidedly with field current. The 
same is true when the field is excited from a separate source. 

The better damping when the motor is operated without field ex- 
citation is explained in part by the greater field-shift or cross-magneti- 
zation for a given load. This results in a greater resistance effect 
relative to the stiffness, the resistance effect being largely due to the 
currents induced in the damping windings. There are other factors 
which give an advantage on the score of damping to the motor which 
has no field excitation. Most of the writers cited in the bibliography 
show both by theory and by test that stability is best at low excita- 
tion, that resistance is prejudicial, and that heavy motor loads ag- 
gravate the tendency to oscillate. 

Positive and Negative Damping Factors. As may be seen from the 
diagrams of Figs. 7 and 8, an increase in generated voltage (at a 
given value of displacement angle) has the effect of decreasing the 
electrical power intake as a motor or of increasing the power output 
as a generator, or, in either case, to cause mechanical retardation. 
Conversely, a reduction in generated voltage favors motor action or 
discourages generator action, and thus helps forward motion. This 

532 S. READ, JR., AND E. W. KELLOGG [J. S. M. P. E. 

is true whether the machine has a large or a small ratio of uL/R, but 
the effect is more pronounced in the latter case. 

Changes in generated voltage may result from either changes in 
speed or changes in field strength. Those due to speed result in 
damping of oscillations, since they cause a retarding torque during 
moments of over-speed and a forward torque during the part of the 
oscillation cycle when the rotor is falling behind in position. Further 
damping is provided by the effect of pole-face grids or damping wind- 
ings which absorb energy whenever the cross-field changes in strength 
or direction. 

An oscillation is accompanied by periodic weakening and strength- 
ening of the field. As was pointed out in discussing Fig. 8, a back- 
ward displacement of the rotor causes increased power intake (and 
this provides stiffness, which is desirable) and it also causes a demag- 
netizing component of current to flow, while a forward displacement 
results in a magnetizing current. The fluctuations in field-strength 
that these magnetizing and demagnetizing currents produce would 
only help to increase the stiffness of the machine, were they exactly 
in phase with the forward and backward displacements; but because 
of its stored energy the field-strength actually lags considerably be- 
hind the factors tending to strengthen or weaken it. As a result of 
the lag, the field is weakest just after the greatest backward deflec- 
tion, and strongest just after the greatest forward deflection. Re- 
membering that a reduction in field-strength tends to cause motor 
action or forward torque while an increase in field-strength results in 
generator action and a backward torque, we see that the fluctuations 
in torque are so timed as to supply energy to the oscillations, or to 
produce what might be called "negative damping." 

A small armature phase-angle (angle p of Fig. 6, corresponding to 
a small ratio of reactance to resistance) tends to magnify this nega- 
tive damping for two reasons: (1) it increases the magnetizing and 
demagnetizing currents which result from a given deflection, and (2) 
it makes the power intake more sensitive to the magnitude of the 
generated voltage. The second of these factors applies also to the 
positive damping which results from the changes in generated volt- 
age with changes of speed, but the slight benefit which may result 
from increasing the positive damping by resistance is much more than 
offset by the increase in negative damping. Both the positive and 
the negative damping also depend on persistence of field-strength, 
but in different degrees. 


Quantitative analyses of the actions that take place when a syn- 
chronous machine hunts lead to rather complicated formulas, and 
bring out some factors besides the ones enumerated above, but re- 
lated to them, and likewise dependent on the tendency of the field 
magnetism to change only slowly. Some further discussion will be 
found in the Appendix. The negative damping is by no means a 
theoretical factor of minor importance, but may, unless motors are 
well designed and used in the best way, cause serious trouble, and in 
any case causes impairment of the steadiness of the motor. 

A-C-D-C Motors. Motors of this class are essentially the same as 
rotary converters. In some machines, the d-c and a-c windings are 
separate, but wound in the same slots. This does not materially 
alter the general characteristics of the machine and we shall consider 
the single-winding type of machine. 

Since an a-c-d-c machine or rotary converter necessarily has a 
rotating armature and stationary field, and it also has a ready source 
of direct current for exciting the field, all machines of this type are of 
the d-c field type. Placing the a-c winding on the rotor instead of 
on the stator means some sacrifice of available winding space. The 
machine may therefore be expected to have a somewhat higher resis- 
tance than a motor of the variable-reluctance type of comparable 
rating, and this is a disadvantage, but the disadvantage (so far as 
efficiency is concerned) is in part offset by the higher power-factor 
and consequent reduced winding current which a d-c field makes 
possible. In the matter of maximum or pull-out torque, the machine 
with d-c field not only has greater stiffness, but can be displaced far- 
ther before it reaches the point of maximum torque. 

Since the earliest days of parallel operation of synchronous ma- 
chines, rotary converters have been more prone to cause trouble from 
hunting than simple synchronous motors and generators. Note, 
for example, the following quotation from Electrical Machinery, by 
Dawes (M-cGraw-Hill, 1922; 11, p. 374) : 

"The converter is very sensitive to line disturbances such as fluctua- 
tions of voltage or frequency. Accordingly it has a much greater 
tendency to 'hunt' than has the synchronous motor." 

The fact that power can be supplied to and given out from a rotary 
converter through either the a-c or d-c connections makes it necessary 
to control several factors which cause no trouble in simple syn- 
chronous motors and generators. Thus, it is quite easy to cause a 
machine to take in power from the d-c side and feed power back into 

534 S. READ, JR., AND E. W. KELLOGG [J. S. M. p. E. 

the a-c line while other machines with which the first is in parallel are 
doing the reverse. If the armature and d-c supply are of low resis- 
tance, the d-c power intake is sensitive to excitation or generated 
voltage, since (unlike a d-c shunt motor) it can not change speed to 
compensate for a change of field-strength. In any rotary converter 
(and to some extent in machines with double windings) the a-c and 
d-c voltages bear an almost fixed relationship. 

From the standpoint of armature oscillations, the same factors 
which produce both positive and negative damping in simple syn- 
chronous motors are exaggerated in the case of a rotary converter or 
a-c d-c motor, for the reason that in addition to the changes in power 
intake from the a-c side there is the additional and far more sensitive 
source of power flow through the d-c terminals. The direct current 
which flows through the armature is equal to the difference between 
the voltage applied to the brushes and the counter-electromotive 
force, or generated voltage, divided by the armature resistance. 
This voltage difference (in an efficient machine) is quite small in 
comparison with the applied and generated voltages themselves. In 
other words, a rather sensitive balance determines the power intake. 
In tests of the small a-c d-c laboratory motor previously mentioned, 
connecting the brushes to a constant- voltage d-c supply increased the 
positive damping more than it did the negative, damping, making the 
machine more stable and also stiffer. This was true provided the d-c 
line voltage was equal to or higher than the brush voltage before 
closing the switch. Material induction of the d-c line- voltage made 
the machine unstable. Low d-c line-voltage makes the machine put 
out d-c power and take in a-c power. It has already been pointed 
out that stability becomes poorer with high motor load. The fact 
that stability was critical to this voltage relation is an illustration of 
the fact that a-c-d-c motors require more careful control of operating 
conditions than simple a-c motors. 

If an a-c-d-c machine is operated from the a-c side alone, but its 
field is excited from its own d-c brush voltage, there is a cumulative 
effect of armature reaction. Thus when the armature reaction 
weakens the field, the d-c voltage falls and this in turn reduces the d-c 
excitation. In this case, the time lag between the field-strength and 
the armature reaction which originally caused it is increased, and the 
magnitude of the field-strength fluctuation is also increased, with the 
result that the negative damping or the tendency to oscillate is ag- 
gravated. The magnitude of the effect just described would be ex- 

May, 1941] 



FIG. 12 (a). A-c d-c motor; no d-c connection -Enng = 69 
separately excited // = 0.390; no resistance in a-c leads. 


FIG. 12 (6). A-c-d-c motor tied to 103-v d-c Enng = 69 
I/ = 0.390; no resistance in a-c leads. 

FIG. 12 (c). A-c-d-c motor not tied to d-c -Enng = 78 
separately excited // = 0.390, R = 3 3 A ohms. 

FIG. 12 (d). A-c-d-c motor, tied to 117-v d-c ri n g = 78, 
If = 0.390, R = 3 3 A ohms in a-c leads. 

536 S. READ, JR., AND E. W. KELLOGG IJ. S. M. P. E. 

pected to depend upon the natural frequency of the motor as de- 
termined by the motor stiffness and the total inertia of the motor and 
the l